Getting closer to trying this on the air.  Just waiting for a resistor to arrive and I'll put it all together.  Hopefully this weekend or the next.

Okay so I'm still not quite ready to try this amp on the air.  My shack is in a "man cave"/storage room/spare bedroom that was a serious cluttered mess.  I spent the last 2 weekends de-cluttering and it's down to about 30% of what it was.  I still have more stuff to move out to the garage and attic, but I needed room to set up this amp, and I wanted to set up a proper table for the radios, my laptop, and a TV.  I am still trying to work out where the big antenna tuner and the amp will go.  I might get a tool cart from Harbor Freight or something to hold it.  It will definitely make moving that big heavy beast around a lot easier if it has wheels.

Due to the way my house is wired, it is simply impossible to have a dedicated circuit for the amp.  I might be able to arrange a dedicated outlet, but it will be on an existing circuit.  If everything works and it turns out the amp is an amazing, huge game changer, I might have a way to run a 240V line for it, with a lot of work involved.

I've got a glitch resistor and fuse for the HV circuit and will install those as well.

This whole project has been a long time coming.  I was given the amp several years ago, and brought it inside and began cleaning and disassembly at the end of last summer.  I didn't really dig into it until a couple of months ago.  So I guess a few more weeks won't make a big difference.

Attached File


You will have a shack heat for the winter....I like the Harbor Freight cart idea for the amp and tuner.

I talked to a ham at a big hamfest that made a plug that plugs into the 240 v 40 amp dedicated electric clothes drier circuit that is connected to 60 amp 240 v romex that snakes it's way to the shack where there is a surface mounted receptacle on a work bench or something and when this ham that I do not know wants to play radio with his amp he apparently unplugs the drier, plugs in his makeshift cord and runs his amp on 240v.

Of course I told him that the whole thing is not to code and he said something like he never leaves it plugged in and it can easily be removed and is no different than an extension chord and that it is just temporary until he pays an electrician to make it all good, but he would require a new circuit breaker panel since his is already full and probably an upgrade to his 100 amp service. Not sure I got all the details correct, but I would totally not recommend that though since it is not to code. He did mention it was a game changer though.

Haha yeah I have a plan...

My drier is way on the other side of the house in the garage, so running a cable from there wouldn't work.  But the previous owner had a jet tub installed in the master bathroom.  There is a breaker panel in the hall closet on the other side of the wall from my shack room.  I think this panel was installed for the tub and might have 240V present.  If so, it would be trivial to pop through the wall with a 240V outlet inside the closet of the shack room and run an extension that way.

That hard part will be getting the wife's Christmas decorations out of the hall closet to do the work.

Okay, so I got impatient about putting in a fuse or glitch resistor and decided to just give the amp a try without them.  They aren't part of the original design, and I wasn't planning to run it hard, just test it out and see if it will amplify a signal and work on the air.

I rearranged my station desktop to get my Palstar AT2K set up.  The built-in coax switch in this tuner is very handy.  It so happens I have 2 antennas, and the tuner has 3 outputs, plus a bypass output that I have connected to a dummy load.

Following Heathkit's instruction manual, I set the amp controls for 20 meters and CW/TUNE and set my radio for 5W output.  I got this working and tuned up, then increased power to 15W and in CW mode intonthe dummy load, it looked like the amp was putting out perhaps 200 or 300W CW based on the meter in the tuner.

I went ahead and answered some POTA activator and they got me in the log.  So it works!!!

But, it isn't without problems.  On 40 meters, the input SWR was over 3:1 as measured by the radio.  I figure it could be several possible problems.

The bandswitch could be the issue, but it works okay on 20, 15 and 10 meters, showing increased power out vs power in, and the worst input SWR was on 10m at 1.5:1 (outside of 40 meters).

So I think it probably isn't the bandswitch.  That leaves the coil and capacitors for 40 meters.  I don't think anyone has changed the tuning slug in the coil, though I suppose anything is possible.  But it could be a bad capacitor in the 40m input circuit, or an open or short on one of the capacitor leads, or possibly the same on the 40m bandswitch leads.  

So, are there any non-invasive diagnostics I can do using my Nano VNA, or my Rig Expert analyzer, or even just a multimeter without having to open the cabinet?

I feel like I am so close to having a working amp, but this one little issue is in the way.  

Anyhow I'd say overall this is a big net positive.  That amp works, it turned 15 watts into over 100 watts, so it's already doing better than my radio's max output, and nothing smoked.  I made a couple of POTA contacts using the amp on 20 meters, so the tubes passed the signal through no problem.  I just have to get the 40m input circuit fixed.

@Jambalaya
Sounds like the amp is working. The Heath manual will have the correct procedure for setting the slugs. Regarding the input circuit,  use the rig expert to look into the input port on 40M, use the same length of coax as the piece that goes between the rig and amp. If you can get a reading that matches the 3:1 SWR, try the slug gently in case it's stuck, you don't want to break it. The alternative is to use the rig to supply 10 watts or so and adjust the slug for the lowest SWR reading on that band. Check the Heath Kit procedure, it can probably be blended into use with the more modern equipment. hth
73,
Rob

I found this, which looks promising because access to the tuned circuits requires the amp to be opened and partially disassembled and powered on to check input.

But according to this, you can leave it off and unplugged and close the T/R relay manually to analyze the input circuit.


Hi Skip,
I have used my VNA to analyze the low pass filter / matching networks on my two restored SB220's by connecting the vna port 1 to the RF in, and connecting the vna port 2 to the cathode terminals to grid underneath with a coax with clips at the end.  
Port 2 is 50ohm impedance, or you can put a 50ohm resistor to simulate the tubes nominal input but you lose the filter attenuation curve.
You have to close the TR relay with your finger to get the plot.
Tubes vary a LOT with drive, so this is just a simulation/emulation.
I used this method to track down an unsoldered capacitor in the 40m network on one of my amps obtained at a swap meet.
This would be a method to check the filter curve.
I can email you the VNA plots from my known good amps you could compare to yours, my email is on QRZ.

IMO one should NOT adjust these inductors unless a previous owner has messed them up.  They are not a tank circuit.
The inductors are set at the factory and glued in place.

Here is the link to a technical explanation actual purpose of the "tuned" input circuits, and the table mentioned  by K3HZP is near the bottom.
https://www.w8ji.com/tuned_input_circuit.htm

Good luck, I'm very interested in finding out how things turn out on your amp.
73  Jim AI5EG

https://groups.io/g/HeathKit-HFAmps/topic/heathkit_sb_220_and_sb_200/91783668

Thanks for the pointers.  I've looked at the manual and schematics and it looks like there are 2 mica capacitors for the 40m circuit, 400 and 470 pF.

When cleaning the amp, I did remove the coils from the chassis and snap them back in.  There is a fair chance that one of the leads got shorted to the 40m coil while removing or reinstalling it.

I'll need to take the front panel off and see if I can see anything obviously wrong.  If so, and I make any changes, I can use the test procedure above to check it with my analyzer before putting like 30 screws back in.

Well, I took the front panel off and looked at the input network.  The 40m coil still has some wax or glue in it from the factory, so I think the slug being out of tune can be ruled out.  Both capacitors looked okay on either side of the coil.  I didn't remove them from the circuit and test them individually.  One of the leads coming from the coil to the back of the switch wafer might be pinched or shorted by one of the coax cables coming in through the chassis floor.  I tried prodding here and there with pliers but it didn't have any effect on the SWR.

So I am down to one of the caps has failed internally (I've read these silver mica caps can change value due to silver dendrites forming inside over time).  Or alternatively, it could be the issue I mentioned with the coax pushing on the back of the switch.

I did rotate the switch and observed the 40m contacts deflecting a little when the switch passed the 40m position, so I don't think it's a problem with the contacts.  I really think it's got to be either the wiring behind the switch (which will be a terrible PITA to get to) or the caps have gone bad.

I could de-solder the caps and test them.  This will be difficult, but still easier than getting at the back of the switch.  If one or both test bad, I can order replacements, re-assemble and move on.  If they test good though, that complicates things.

You might try driving it a little harder, I have run into these that have higher SWR at lower driver than they do at 80+watts of drive.  The input circuit on the SB isn't normally that good, it was made in an era of tube radios that could match a 3:1 input without to much issue, anything less than 2:1 is pretty good for a SB.

If the slug can be moved, I'd gently try it, the case may be that the coil only needs a bit of adjustment. This is lots less trouble than checking components. hth
73,
Rob

Yes I would attempt to adjust the slugs. If you can get it below 2:1 then most of the rigs will not fold back power. I was able to het most of my SB-200 below 1.5:1. Remember these were designed to be run with a tube radio so the swr was not tuned flat. A soldering iron can sometimes help to free a stuck sluck. Use a non inductive tool to adjust them otherwise you will get false readings. Also be careful to not crack them.

If the amp had never been used with a solid state radio, that maybe how it was built. Most of these amps were kits and there was no factory involved. A tube radio has a lot more leeway on the input impedance so it may have just been like that since being built a 3:1 SWR is generally considered what a tube final can match just by tuning the final for max output with a stand alone matching network ( tuner ).

I know people frown on using a tuner on the amp input, but I have done it with the AL-80B because to use 30 meters and other bands because 30 and 20 share the same input. You could try that.

Also old caps can leak, which will change the impedance, so a cap can be leaky without any obvious damage changing the value outside the tolerance and creating a different impedance. It might be worth just buying new caps and replacing them if cleaning the band switch contacts doesnt work., if you can get to them.

Also you don't have to get your hands on the back of the band switch to clean it. All you have to do is use a deox contact cleaner with the very thin tube and get that tube down to the contacts and spray it. Don't do it inside. Do it outside and by up wind, that stuff is very toxic but it de-oxidizes the contacts, cleans the carbon and evaporates in seconds. You can position the band switch inbetween bands to break contacts from touching and spray the hell out of the band switch to clean it. I would try that first because it should be very easy to do.

Something like this.

Thanks for all the tips.  I took the top off the amp again today and had another look at it.  I managed to lose the little red tube on my contact cleaner.  I tried the tubes from canned air, WD40, brake cleaner and cutting fluid and none of them fit.

I did clamp a q-tip in some hemostats and dab at the contacts.  I had sprayed them down pretty good back when I had the amp fully apart.  I don't think it's the contacts.

Anyhow with regard to the SWR thing, it seems to increase a bit with increased power.  At 5W it's below 3:1 SWR and at 15W it's above it.   It seems odd that the SWR would change with the power applied.  The other bands are all pretty low, under 1.3:1 except for 10m which is about 1.5:1 and I can live with that.

I didn't really want to try higher power yet until I was sure the tubes weren't going to fail.  Plus I don't have a dedicated circuit to plug the amp into.  I don't even have a dedicated outlet.

I did order some alignment tools and some replacement mica caps so when they arrive I'll give it a try.  

My multimeter will measure capacitance but it only has a nanofarad range and doesnt go down to picofarads.   So a 470pf cap would read 0.470nf.  I measured some cheap HV ceramic caps (so cheap they don't even have a tolerance listed anywhere on them or their container) and they were off a bit as measured, but they were all reading a little lower than their rating so it's probably the meter or something.  

It's a huge pain to get the front panel off removing all the knobs and switches, and then I'll have to do that trick with the 50 ohm resistor and manually actuate the relay in order to be able to tune the slug if I don't want to do it under power.

Adding power will find lots of faults in your system. Just like ft8.

Haha, I'm sure it will!  Just hoping to get everything in place for minimum drama when the faults become apparent.  

If something in the amp goes bang, it's guaranteed my wife will think I had an ND or something.

Yea with how the strip line and diode detectors work it is not uncommon for SWR to change a bit with power increase, especially with a larger amount of power being applied. Keep in mind that many radios are not perfect when it comes to reading reflected power either. I have seen issues where SWR will read 1.4:1 with 5 watts going out, then jump to 2.5:1 with 100w. Even with relatively low SWR on my SB-200 with 20w of drive when tuning, it will jump when I go up to 75 watts. Like I said if you can get it below 2:1 most modern radios will output nearly full power. So that will at least get your drive level up. Many of the tuning slugs on my SB-200 will only tune to a 1.5:1 and some are dead flat.

Hope that helps. If you have any issues with dirty relays check out W8JI webpage on cleaning them with DC voltage. Using deoxit will only help for a couple of days most of the time.

@Jambalaya
IDK if somebody suggested this already, but it's OK to use the internal tuner in your rig to match up to the amp input circuit at least until the input problem is resolved. This depends a bit upon the drive level needed and the duty cycle of the transmitter. IOW, I wouldn't run the rig + tuner very hard into much of a SWR mismatch. HTH
73,
Rob

I read somewhere that using an internal tuner into an HF amplifier can cause oscillations on the tube amp. But I will be damned if I can find it.

ETA: This is from W8JI, if you do not know who he is you should look into it. He says it is ok for the most part. But there is a lot of good information about the tuned input on cathode-driven tube amplifiers.

https://www.w8ji.com/tuned_input_circuit.htm

ETA: It may have been somewhere where they were talking about doing away with the tuned input and using a tuner between the amp and rig as a tuned input. I digress, problem when you are always researching subjects while going down a rabbit hole at midnight lol

Yeah I think I read that W8JI was an engineer that designed some of the Heathkit stuff, including the SB220.  

Anyhow I knew you could use a tuner between the rig and the amp.  But I just want to get it back functioning as it should so that I can use it as intended.

I do have a tuner between the amp and the antenna, a palstar at2k.

There is a little bit of a mismatch on 10 meters too, about 1.5:1, but I've read that's very common so I'm not going to mess with 10m.

I ordered some capacitors and I'll replace them when they arrive and then see if that fixes it.

Yea my LDG AT-600 Pro does not match my doublet perfect on everything. It gets it close. I put the amp in bypass mode and hit the tune button on the Icom and tune the antenna, I then tune the AMP into a dummy load for the band I am on an send it. Works great. You can touch it up on the air but the power difference is so minuscule no one would know the difference. One of the reasons this new LDMOS HF amp I am building will have an auto tuner built in while it is a single pallet unit, but once I go to dual pallet I will need an offboard tuner.

Yeah that's pretty much my procedure.  Put the amp in bypass (by unplugging the relay cable) and tune the palstar to the frequency using my radio, then switch to dummy load and tune the amp for the band, then switch back to the antenna.

That amp you're building has to be a fun project.  I would love to build something like that in the future.  It's not the cost of the amp that's prohibitive so much as the cost of entertainment to get the wife and kids out of the house so I can actually build it.

My amp and tuner are all in one. I turn off the amp, tune the band, and then forget to turn the amp on, and transmit with 5w for a while.

Hehe I'm sure I'll do something similar.

My biggest worry is forgetting to change bands on the amp or forgetting to switch to the dummy load when tuning, etc.

I got some replacement mica caps in today.  

They were advertised as I have labeled onnthe paper sheet in the picture below.  But they are considerably smaller in size than the original caps they will replace.  

The voltage ratings are not printed on them so I am hoping it's just a case of modern manufacturing and miniaturization making them smaller than they were 50 years ago.  

I'm going to replace them this evening, hopefully.

Attached File

So I think I'll call this a success.  I measured the capacitors after pulling them out and the 400pf cap measured 407 on my meter, and the new cap was like 417, so I kept the old one.

The 470pf cap measured 436, and the new one was like 477 so I used it.  The leads were a smaller gauge of wire and I hope that doesn't present a problem later on.

I also poked at the goop holding the tuning slug in place with an Allen wrench and it is like a soft, non-adhesive wax that probably got a bit of heat from all the soldering in the vicinity.  I think if I ever want to turn it, I'll be able to.

For now, I am able to get approximately 600W (using the 3000w switch on the tuner) into a dummy load on 40m with a little over 2:1 SWR.  I am satisfied that this will probably be okay for SSB, and also I think it will be low enough to use the rig's auto tuner to touch it up.

Attached File


Multiply this reading x10.

Attached File


Attached File


Prior to replacing the capacitor, I was getting about 3.5:1 with 15 watts out of the radio.  This is with 40 watts out of the radio.

Keep in mind modern capacitors are usually much smaller than the ones from those eras. Those caps based on the size are probably 500v rated. I have some 22pf like that for my AM-6154 conversion that are rated for 2kv so they could be higher. On the input side they are probably more than sufficient.  If you are getting 600w with 45 watts drive it sounds about right. With 45 watts drive I get about 250-300 out of my SB-200. And about 100w with 10w of drive. Not knowing the condition of the tubes its hard to say. But I would adjust the slug a bit more. Check some of the Antique radio pages for a slug tool. They are double sided non inductive tool that has a the common hex sized. This way you take less of a chance of cracking a slug and getting it jammed in the tube. If this happens its not a good day. Been down that road. And trying to get a ferrite slug that is close to the same size and type in nearly impossible .

The non inductive tool allows you to make the adjustment as you transmit and dial it in without a metal tool messing up the inductance  and yes the soldering iron trick works great. Used it on all the old radios I use to restore. I always use the non inductive tool and see of it will move, if it does great. If not out come the iron.

ETA: about the slugs, a lot of these radios and amps have different size slugs for each band. So if you crack one you can not borrow one from another band that you do not use much.

Yeah I will keep all of that under advisement.  I didn't try to turn the slug, I just poked an Allen wrench into the hole to clear it.  The goop has the color and consistency of dental wax.  I think it will turn, or if not, a heat gun will soften it enough.  It's already pretty soft at 80 degrees ambient temp in my shack.

ETA I do have some plastic alignment tools somewhere in a box but I probably will get new one in the mail before I find the ones I have.

@Mach

Going back through this post, where he says "connected VNA port 2 to the cathode terminals to grid" is confusing me.

My grid terminals are connected directly to chassis ground and I'm guessing the cathode is the filament in this case, but the filaments have 2 terminals each.

So if I had to guess, I would be connecting a 50 ohm resistor between the 2 filament terminals where they connect to the filament choke and the other side of the resistor to ground?  

Then just jumper the relay contacts and I can use an analyzer connected to the coax input to tune the slug?

ETA I think I might also need to remove the tubes for this procedure.

I think it was around 40W on 40m and 50W on 20m.

Reading it again I dont know what he means. I think you should be able to close the relay mechanically and set the band switch to 40m and look at the input circuit. I have not adjusted the input circuit on the 220, but when I adjusted the slugs on the AL-80B, I just did it by the SWR meter on the radio, but that amp has easy access on the back.

I think if you just read the SWR at the RF input coax connection, with the tubes in but amp not plugged in and close   the T/R relay manually you should get an SWR curve on the S1 and adjust the slug  but I would test or replace the 2 caps on the input circuit before adjusting the slug, but you could just try adjusting the slug.

Okay so minor update.  I was digging around in my totes of electronics parts for some resistors today and couldn't find them.  I'll have to go back through them again more thoroughly.  I am tempted to just use a potentiometer set to 50 ohms because I know where one of those is.

But I did get some latching switches and tiny project boxes yesterday.  I split my amp keying cable in two and put a latching push button switch in.  This allows me to interrupt the keying cable without having to reach behind the amp and unplug it.  So it's basically a standby switch that will let me pass through the amp so I can tune the antenna tuner.

The little case needs some superglue, but I used some clear heat shrink until I can find the superglue.

Attached File

So I went to a lot of trouble to try to fix the 40m input impedance and it was a bust.  

I wired 50 Ohms between filament (cathode) and ground.

Attached File


Then I jumpered the relay contacts so I wouldn't have to hold it by hand.

Attached File


I ended up backing the slug out all the way.  Soldering iron heat was enough to loosen the slug.  None of my tools were a fit so I used a shaved down bamboo chopstick.

Attached File


A few successive plots as I backed the slug out.  SWR got better by a little.  But it seems to be actually resonant down around 6.5 MHz.

Attached File


Still, it looked promising having come down from over 2:1 mismatch down to 1.5:1.  

Unfortunately, after putting everything back together and applying power, I ended up with over 3:1 SWR.  

So, both mica caps measured okay on my meter.  I tried the coil at many different positions and got it down to 1.5:1.  But something is changing the match when power is applied.  I wonder if it's the other cap mica cap (the one I didn't replace) failing?

Anyhow that is a few hours of work for a disappointing result.  I am not going to put any more money into this amp.  I might try to track down the problem again, but I need to take a break from it.  I can't find whatever it is that is changing value with power applied.

I've been through the schematic and the drawings in the Heathkit manual and the mica caps are the right values connected to the right parts of the coil, and the right switch contacts.

W8JI designed all the tube Ameritron amps and at least some of the Healthkit amps.

I had a short email conversation with him when I had to fix my AL-80B. He gave me a list of components to check based on symptoms.

He is a very nice guy to help a rando ham out after buying a used amp he designed years ago.

I have seen a few of these that the input SWR gets better with more drive.  The current on I have posted pictures of is over 2:1 on 80 with 10 watts drive and less than 2:1 with 100 watts of drive.  Have always tuned the inputs on the SB's with a radio and dummy load while in the CW position, kinda crude, but it seems to work.  Just have to be very careful not to short anything while the amp is operating while half tore apart.

I don't know if this is happening in your case, and I don't know if I am fully remembering it correctly, ( tried to google it but didn't get anywhere ) but an old cap can have increased insulation resistance and decreased breakdown voltage.

LCR meters measure the capacitance based on the time it takes to fully 'charge' the capacitor with a dc current. It isn't really measuring capacitance, but measuring time at a low DC voltage and known current until the current stops flowing into the cap. Of course that is an indication of capacitance, but, an old capacitance can develop an internal resistance and give a false reading of higher capacitance.

Also a cap with age can have a decreased breakdown voltage. The LCR meter doesn't measure this either. As voltage increases, capacitance goes down because it leaks current instead of accumulating a charge.

I think it is possible, especially with electrolytic and paper, to have both and the LCR meter indicate a good capacitor but when the AC voltage goes up, the capacitance goes down and  changes the impedance of the tuned circuit and create a higher SWR.

I do not know what the likelihood of this happening in your case, but you changed one of the caps with a new cap and the SWR went down from 3.5 to 2.0. You alreadu have the other new cap. If it was me, I replace the other old cap with the new one.


--------------------------------------------------------------------------------------------------------------------------------------------

I have found this from W8Jii and it explains what that post was talking about and he gives the value of 100 ohms for a 3-500Z tube as the impedance of the tube as seen by the tuned input circuit when the tube is at full drive.

https://www.w8ji.com/tuned_input_circuit.htm


Input Impedance

The driving impedance of a cathode driven amplifier always varies some amount with power level and tuning. We can say the input impedance of a cathode driven stage is dynamic, with the tube's driving impedance changing with parameters like drive power and how the amplifier is loaded or tuned.

The input impedance changes dynamically because the output system is in series with the input system. The same plate current flows through both cathode and anode systems. Varying anode load impedance, because the same very same current flows through both, changes input impedance. In general the lower tube mu, the greater the effect of output tuning and power level on driving impedance. This is because a low-mu tube generally has a high driving impedance, making cathode impedance a larger percentage of anode impedance. The higher cathode impedance increases negative feedback, and increased negative feedback causes the input impedance to vary more with anode load impedance changes. The tuned input's resonance or "flywheel effect" does not stabilize this impedance variation, but additional losses brought into the cathode system by tuned input circuit components helps dilute or "swamp out" dynamic changes in driving impedance. An attenuator pad would do the same of course, but would not filter harmonics or sharpen conduction point transitions.

The dynamic cathode impedance, where cathode impedance changes with drive level, prevents us from accurately adjusting the tuned input system of a typical cathode driven amplifier at very low power. This prevents accurate use of an antenna analyzer to adjust the tuned input while using the operating PA tube as a load. The input system should always be adjusted near full output power under normal operating conditions, or with a suitable dummy load replacing the tube's dynamic impedance.

A reasonable approximation of tube input impedance, for circuit adjustment purposes, would be a small non-inductive resistor from each cathode pin to a control grid pin at each tube. This resistor should be the approximate value of the tube input resistance when driven near full power with normal tuning and loading. Filament power and HV should be off and the tubes left in place, and the input driven normally as if the amplifier was running but with a low power antenna analyzer substituting for the the normal exciter.

For various tubes the following cathode resistances are used at each tube:
811A, 572B 220 ohms
3-500Z 100 ohms
4-1000A 90 ohms
3-1000Z, 3CX1200

3CX800A7
68 ohms


I think this will be your key to solving this. Unlike an amp with access to the tuning slugs on the back, the SB-220 slugs do not have easy access. So adjusting the slug in between high power key down would be very time consuming at best.

First I would replace that other tuned circuit cap.
then I would test the amp SWR when operating on 40 meters at a higher drive level
then if the problem is not solved ( but I think it might be ) I would do the resistor dummy load across the tube and adjust the slug with the meter.

Yeah I think it's possible the other cap is bad.  It's one of the only plausible explanations.  It's just odd that the other bands are all okay.

Unfortunately now I'll need to tune the slug regardless because I moved it.

I used a 50 ohm resistor (two 100 ohm resistors in parallel).  But there are 2 tubes in parallel so 50 ohms is correct in this case.  And I checked the other bands while this was set up and they all read low as expected.

Link is not working.

Here is a high quality set of tools for tuning that may come in handy. Also DX engineering has the ameritorn tuning tool for about 7 bucks, and may work as well.

It is curious that only the caps in the 40m input circuit would be bad, and apparently both of them. Which seems odd.

But the bottom line is you had 3.5 SWR and changed 1 out of 2 caps and the SWR ( and therefore the impedance ) of the input circuit got better. If the cap you took out was perfectly fine, nothing would have changed, assuming the solder connection was good, which could have been an issue maybe.

Let me try to connect some dots.

1 slug and 2 caps     SWR was 3.5

Replace 1 cap     SWR 2.0

Back the slug out all the way   SWR 1.5, with better SWR below the band, resonance say around 6.5 Mhz.

Some clues:

With an inductance and capacitance tuned circuit

When you raise the inductance, you lower the frequency.
When you decrease the inductance you raise the frequency

When you raise capacitance you raise the frequency
When you lower the capacitance you lower the frequency

You replaced a capacitor and the SWR got better ( 2.0, was 3.5 ) with lower SWR lower in freq. So you increased the resonant freq which means you increased the capacitance when you replaced the capacitor.

Then you backed the slug all the way out, decreasing the inductance, and raising the resonant frequency improving the SWR to 1.5 with lower SWR still at a lower freq..

If you were to put the slug back to where it was, you would be adding inductance which will lower the frequency, which is the wrong way to improve SWR on the 40m band.

It appears that the tuned input circuit does not have enough capacitance to get the lowest SWR in the 40m band.

You need to add capacitance. You did this once when replacing the first capacitor and the SWR came down, with lower SWR at lower frequencies, which means 2 things.
You increased the capacitance, which means the capacitance of that old capacitor was much lower than spec and probably much lower than read by the LCR meter.

When you put the slug back to where it was, you will be increasing inductance which will lower the freq and put the SWR back to 2.0
You will need to add even more capacitance.

So it really looks like the other cap is indeed bad with a very low capacitance value.

Remember an LCR meter does not read capacitance. It estimates capacitance of a working capacitor by using a known DC voltage and known DC current and measures the time it takes to saturate the charge ( current stops ) and then calculating what that capacitance is but in no way actually measures the capacitance when used with AC / RF.  Old caps can have the dielectric break down reducing the actual capacitance and the DC charge it can hold AND also develop an internal resistance. So you can end up with a bad cap that the charge it can hold is very much reduced ( lower capacitance ) but the time it takes to charge that cap is very much increased so the very bad cap on an LCR meter looks good. Ie  very low capacitance due to dielectric breakdown but due to internal resistance, takes longer to charge like a higher capacitor would take ,

I think there is a good chance that is what is happening.

But I could be wrong.

So I took the amp apart again today and soldered in the other new mica cap.  The impedance didn't change.  In my previous posts I was just guessing at the resonant point because I hadn't swept it with a wide enough range.  Today after swapping in the other replacement cap and seeing no change, I widened the sweep and this is what I got.  So it looks like it's resonant at 5.723 MHz.

Attached File


Just as you said, moving the slug in moved the resonant frequency farther down. (The above sweep was with the slug all the way out.)

So next I began to wonder if perhaps the wrong coil was installed in the 40m position.    An image search of SB220 amp pictures online seems to indicate that it's the correct coil, presumably with the correct inductance.

So, that means something else is affecting the impedance since I've eliminated the coil and the capacitors.  And, whatever it is, only appears to affect 40 meters and not the other bands.

It's a real head scratcher and the only thing I can think of is maybe there is something wrong with the band switch.  But I inspected it pretty carefully.  All 4 contacts on both sides of the wafer deflect when it is switched into 40 meters, so I know they are making physical contact with the central rotor part of the switch.  There is a bright silver line there as well, and I have applied copious amounts of contact cleaner and cycled it many times.  

So, this time I didn't bother to reassemble the amp, instead opting to try further diagnostics when I have more time.  I think I can make some short alligator clip jumpers and jumper the band switch contacts, taking the switch out of the circuit.  If the problem persists, I will know it isn't the band switch.  

I've looked over the schematic lots of times and I don't see any other capacitors that could contribute to the problem that wouldn't also affect the other bands.

I also don't see anything in series that would add inductance.  It's also very strange that after replacing the mica capacitors, it still exhibits the same behavior where an increase in exciter power causes a higher SWR.  This doesn't appear to happen with the other bands, only 40 meters.  

Also there is a greater than zero chance it's a problem with my radio, perhaps an ATU latching relay stuck or something.  I probably should try driving the amp with a different radio as a test.  My radio's internal tuner has been turned off and bypassed for all of this, but you never know.  It's not likely though considering the SWR plot showing resonance way down at 5.723 MHz. Also I can check the radio itself.  If it reads low SWR into a 50 ohm dummy load on 40m then it's probably not the radio.

But I just don't see anything else on the input side of the amp that would only affect 40 meters.

very surprising the cap didn't fix it.

Based on other pics, the inductor for 40 meters should be the one with the second most number of turns.

the 80 meter slug should have the most number of turns, the 40 meter slug the second most number of turns and so on up in freq.

I too have looked at the schematic looking for something else that would only influence the 40 meter band. I don't see anything either.

I am assuming you got that graph with the 50 ohm non-inductive resistor across the tube terminals to produce a 50 ohm load to the input circuit right?
If so i don't think the rear wafer switch that selects the inductance for the load selection of the band comes into play.

I will take another look though, because I like trying to solve mysteries with equipment.

If you put the slug  back to about where it was, how low does the resonant freq go?

Could this be an 80 meter slug.

It looks like your 40m input is now a 60 meter input.

One side of both of those caps go to ground.

have you checked the continuity from where the caps connect to ground to actual case ground?

a poor ground connection would change to capacitance or even remove it completely from the circuit.

The cap ground connection is that aluminum screw attached to the aluminum case, maybe there is corrosion between the screw head and the case causing an RF impedance but I would think if the head of that screw has continuity to the case it would be good. the entire build is made of screws into the case for ground connections so I don't think it would be likely, just grasping at straws. If it was a poor connection though I think it would add RF capacitance, not take it away.

I think I have been thinking about this wrong.

The filter is a low pass pi filter, it has an upper cutoff freq and an impedance

But using this calculator

https://3roam.com/pi-filter-design-calculator/

doesnt give a 50 ohm impedance for those values on the schematic.

So I am confused on what that tuned input is actually doing.

In the case of RF applications they are used as an impedance match before an RF amplifier, but I am having trouble understanding exactly how.

I am missing something

I think the impedance changes not like the resonance of a simple LC circuit.

If you put the slug back to as close to the original position what does the SWR do?


With the resistor set up for a 50 ohm impedance as seen by the input circuits do you get good SWR on the other bands?

I think it is possible there is a mistake in where the 2x 100 resistors are being placed for the dummy load....maybe.

If the other bands are good, then maybe not.

Ok reading it all again.

I think you need a non-inductive 100 ohm resistor from each of the 2 cathode pins to a grounded grid pin, and that done for each tube.

It looks like you only have 2x 100 ohm resistors ( equivalent of a 50 ohm resistor )in parallel  from only one side of the connected cathode pins to a grounded grid pin only using one side of the cathode pins connected together of both tubes.

I think you need another 50 ohm resistor connecting the other side of both cathodes to a grounded grid pin.

Am I correct or am I reading incorrectly about what you did.

Ok I think I figured out the problem

I think you need another resistor from the other cathode pair to ground, otherwise the impedance you have is the 50 ohm resistor in parallel with the impedance of both cathodes and you do not have a 50 ohm load.

As described it is a 100 ohm resistor from each cathode pin of each tube to ground and for a single tube that would be it but with 2 tubes that have each cathode pin connected to the other cathode pin of the other tube, a 50 ohm resistor on each common cathode to ground would be the same thing since the cathodes connection of the tubes are in parallel.



disregard this is all wrong. The basic short through the cathodes is just a loop of wire that has no other connection, you still have a 50 ohm resistive load on the output of the tuned circuit.

I am an idiot. back to more coffee. I am out for a while

All this thinkin hurts

Here's the thing........if you are actually worried  about   high voltage leakage/ arcs/  faults in  your high voltage transformer,  you cannot  conclusively  test for that  AT LOWER VOLTAGE.   You need to use HIGHER voltage   than the secondary sees,  to actually meaningfully  test the thing.  Not even  a  "hipot"  or megger  can test for that.

Probably, if you are worried about the transformer,  probably  the only way  to test it is something on this order:

Take the max voltage the transformer sees, either the secondary, AC  or the max  DC output,  and multiply that X 2

Then  come up with  either an AC or DC source at that voltage,  and connect it with metering and current limiting---on an insulated test platform---between one  primary lead and the case,  and  then between one  seconary lead and the case,  and then between one primary lead and one secondary lead,  etc,   until all windings have been  tested against the case and the secondary.

If this is not a sealed transformer,  AND if it has seen ANY  questionable humidity/ storage,  I would first bake it at low temp  for  24 hours,  say,  at least 150 degrees F,  to drive out moisture  It would It would not hurt, if you can come up with a method,  to take the still hot transformer  and evacuate it, say, using an old pressure cooker, whatever.

You DO have a vauum pump, right?

DO NOT  repeat  DO NOT   short out filter caps with  a   shorting stick  FIRST.

DO THIS instead.

Find some old just wirewound resister,  probably at least 2K or more, less than  50K  Make  a shorting stick with the  resistor  up near the probe.  USE THAT FIRST  and WHY???

Because IF   you have a power supply,   AND THE BLEEDER  has failed, or there is none,  you can DAMAGE  or  KILL  filter caps  when  "dead" shorting them,  because of the  peak CURRENT!!!

So START with your  "resistive"  shorting stick,  and when you believe that is worn down,  THEN get the "dead short"  stick.

ALSO ALSO  on such amps as  an SB220, you can  KILL THE METER(s)   with high  discharge current through the caps.