• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Heathkit AA 100

Z

zdenekh

New Member
Hello,
Could someone please advise me? My Heathkit AA-100 was working after I made the following modifications: I replaced the selenium rectifier with a 1N4007 diode and used an 8.2 kΩ dropping resistor. I measured -18 V at the test point after the resistor and approximately -18.2 V at pin 6 of the new Tung-Sol 7591A tubes (with a 10 kΩ dropping resistor, I had measured just over -17 V at pin 6 and -16 V after the resistor). I also replaced the 470 kΩ resistors with the recommended 220 kΩ ones and increased the coupling capacitors to 0.22 µF. The unit was working. Later, I acquired a digital bias meter, which showed a reading of around 70 mA for each 7591 tube. I reduced the current to the required 35 mA by swapping the 8.2 kΩ dropping resistor for a 5.6 kΩ one, but the rectifier tube subsequently failed. Any advice would be greatly appreciated.
Thank you
Zdenek
 
Register to hide this ad
Did you replace the power supply filter caps? Thats often what kills rectifier tubes. The other side of that is they can't be significantly larger than stock, otherwise that will also damage rectifier ubes.

If its a modern rectifier they tend to not be as robust as old ones. Also quite frankly sometimes they just wear out. Exactly how did it go? Spark show and a blown fuse, or did it just drop output voltage?
 
Yes, the filter capacitors are new—from JJ Electronics—and match the original capacitance values. The TAD rectifier tube was also brand new. I replaced the 8.2kΩ dropping resistor with a 5.6kΩ one. The unit was on for about five or seven minutes, and I measured a quiescent current of around 35 mA; everything was fine. Then, while trying to measure the negative voltage on pin 6—with the unit positioned vertically—the rectifier tube sparked, there was a bit of smoke, the left channel started whistling, and the rectifier tube seemed to go dark; I switched it off immediately. I never actually got to measure the negative voltage; I only had the multimeter's black lead connected to the chassis. I'm not sure if the mere replacement of the resistor caused it.
 
Whistling in the left channel and the extinguishing of the rectifier tube occurred after the device was switched on for the second time via a light-bulb tester.
 
well if the rectifier shot sparks its generally done for. Even if its not permanantly welded together, once they arc its far more likely to do so again. Possible internal bits shifted around from standing sideways. That shouldn't be a problem but if it was already jarred loose from shipment it may have just been the final straw.

There is a common mod with Dynaco amps that adds a silicon diode in series with each plate lead to help with PIV limitations in modern rectifiers. Not sure if it would really be needed here but I believe the AA100 does run similar voltages to the ST70 so perhaps not the worst idea.

I doubt changing the resistor in the bias supply had anything to do with it honestly. I think you may have just experienced a mix of poor timing and bad luck.

also you may want to report this thread and request one of the mods move it to the Tube Audio section. Probably will get more input there. I've never worked on an AA100 so maybe someone with specific experience has a thought.
 
Thank you very much for the opinions and recommendations. I did have two silicon diodes connected in series with the rectifier tube (following the Dynaco instructions). I’ll give the tube section a try. I need to take a bit of a break from working on the amplifier for now....
Z.
 
Hello,
Could someone please advise me? My Heathkit AA-100 was working after I made the following modifications: I replaced the selenium rectifier with a 1N4007 diode and used an 8.2 kΩ dropping resistor. I measured -18 V at the test point after the resistor and approximately -18.2 V at pin 6 of the new Tung-Sol 7591A tubes (with a 10 kΩ dropping resistor, I had measured just over -17 V at pin 6 and -16 V after the resistor). I also replaced the 470 kΩ resistors with the recommended 220 kΩ ones and increased the coupling capacitors to 0.22 µF. The unit was working. Later, I acquired a digital bias meter, which showed a reading of around 70 mA for each 7591 tube. I reduced the current to the required 35 mA by swapping the 8.2 kΩ dropping resistor for a 5.6 kΩ one, but the rectifier tube subsequently failed. the filter capacitors are new—from JJ Electronics—and match the original capacitance values. The TAD rectifier tube was also brand new. I replaced the 8.2kΩ dropping resistor with a 5.6kΩ one. The unit was on for about five or seven minutes, and I measured a quiescent current of around 35... The mains fuse is not blown. Whistling in the left channel and the extinguishing of the rectifier tube occurred after the device was switched on for the second time via a light-bulb tester.
Any advice would be greatly appreciated.
Thank you
Zd.
 
zdenekh said:
Whistling in the left channel and the extinguishing of the rectifier tube occurred after the device was switched on for the second time via a light-bulb tester
Click to expand...

Based on the descriptions it is hard to provide good advice. There are some gaps in knowledge. What are your power supply voltages? Is your B+ at 440 V or did it creep up, for example. When did you start using the new TAD 5AR4 and Tung-Sol 7591s in the updating process.

The changes you made to the grid leak resistors and coupling capacitors should not cause problems. Neither should the replacement of the selenium rectifier with a 1N4007 cause these problems you experienced. You verified that you had, as expected, more negative grid bias than specified, so your power tube quiescent current should not have risen. Yet you experienced run-away behavior and measured 140 ma per pair of Tung-sol 7591s AND running the 7591s at >30 watts dissipation each IF your B+ was the specified 440 V. Jim McShane specifically warned against running new manufacture Tung-Sol 7591s at over 15.75 watt. Did you notice any red plating? I expect that the 7591 power tube may have red plated, and hope they are still serviceable. Try to test them before further use.

At a total draw of 280 ma for power tubes alone it definitely exceeded the capability of your TAD 5AR4, at least temporarily. ([How dit you measure the power tube currents?). It is unfortunate but not surprising that the 5AR4 failed. A NOS 5AR4 might have survived the abuse, but new ones often do not. Did you do the “yellow sheet mod” for your modern 5AR4? If not do so for your replacement.

So, what caused this run-away behavior? The whistling you mentioned might indicate oscillation. I doubt the changes in coupling caps and grid leak resistors you made caused it, but I wonder if there were some cold solder joints in the process. All I can think of to suggest is to go over every joint you touched and reflow them. Then remove all tubes, except the (new) rectifier and measure voltages. Basically follow the start-up procedures in the Heathkit construction manual.

Good luck and keep us informed.
 
Thank you very much for the detailed comment. I won't be able to examine and measure the unit until next week. I should just clarify that the problem occurred only after I replaced the dropping resistor (changing it from 8.2 kΩ to 5.6 kΩ). The switch to 5.6 kΩ resulted in a quiescent current of around 35 mA per 7591 tube (totaling 140 mA, rather than the 280 mA seen with the 8.2 kΩ resistor). I measured the quiescent current using an Amptata tester for the 7591A; I’ve attached a reference image. The TAD rectifier tube was set up in "yellow sheet mode"—with two 1N4007 diodes connected in series to each of pins 4 and 6. Despite this, a failure occurred. I didn't notice any red-plating. I’ll let you know once I find out more.
Thanks again.
 

Attachments

  • s-l1600 - 2026-09-21T105414.790.webp
    s-l1600 - 2026-09-21T105414.790.webp
    36 KB · Views: 6
Ah! The Amptata tester makes sense for power tube measurements. I wondered if you perhaps soldered in some sensing resistors to measure the 7591 currents.

I don’t have first hand experience with newly-manufactured 7591 tubes but I believe they may require higher negative bias to run at the same currents as NOS. This is apparently confirmed in your unit where you needed to decrease the 8.2 k resistor to 5.6 k to get the power tube current down to specs. Unfortunately the stock Heathkit AA-100 does not offer a convenient way to adjust bias. Your next set of 7591s may require a different bias (or the matched quad may not be perfectly matched), so the ability to adjust the bias voltage for individual 7591’s might be something to consider. Something like: ST35 Bias Control Upgrade - Dynakit Parts adjusted for handling the right voltage range.

Again, the sequence of events is not clear to me, but the damage was probably done while the unit was sizzling away at 30 W per tube, and not by the subsequent correct quiescent current phase, after the current were brought down to spec. Failure of the rectifier during the last phase was probably the last straw due to earlier damage. But the whistling makes one wonder if there is another remaining issue. Do you have access to an oscilloscope to inspect for oscillations if the problem persists?

What is the voltage and value of the first cap in your power supply? Does your JJ multicap have a 60uF/500V? Note 60 uF is the max a 5AR4 can handle. At start up your first cap may exceed 500V so be suspicious of possible damage here.

Suggestion for how to proceed:
1. Verify the status of your rectifier and 7591s. Test your tubes if you are able to, to ensure the 7591 quad has balanced tubes. Get at least a replacement 5AR4. It is good that you have your “yellow sheet mod” done.
2. Check and reflow all joints of your recently added components. Inspect old connections, as well as traces of the PCBs and the tightness/grip of socket connectors.
3. Connect speakers (or other load) and turn unit on without rectifier; preferable on a variac or dim bulb. Check your AC voltages without the 5AR4. Turn off and plug in the rectifier. Repeat slow turn on and measure power supply DC voltages and compare to schematic. Since unloaded it will be higher than specified on the schematic. If within range turn unit off, let cool.
4. Pllug in tubes including your tested good power tubes (and your Amptata). Bring up again slowly while monitoring your tube currents. Measure your B+ when steady state is reached; if is too high you will have to do something (variac or bucking transformer) to get it close to spec. of 440 V on the 7591 plates. Adjust hum per instruction in the manual.
5. Hopefully all is well and you hear good music when you apply a source. Keep monitoring your tube currents for a while.

Good luck!
 
Thank you very much for analyzing the problem and suggesting the next steps. The sequence of events was as follows:

1. The unit was operating with an 8 kΩ series resistor.

2. Using the Amptata tester, I measured the quiescent current with the 8 kΩ resistor; it was approximately 70 mA per tube.

3. I changed the series resistor from 8.2 kΩ to 5.6 kΩ and measured a value of around 35 mA per tube via the Amptata; I monitored the status for about 10 minutes, and the unit operated normally.

4. I wanted to measure the voltage at pin 6 of the 7591 tubes, so I disconnected the tester and put the 7591 tubes back in place.

5. I positioned the unit vertically (on its side), connected the multimeter's common lead (COM) to the chassis, and turned the unit on. At that moment, the rectifier tube began to show signs of trouble—it appeared that sparks were jumping inside (it seemed to me). I immediately turned the unit off.

6. I turned the unit on again using a light bulb as a current limiter; it did not indicate a direct short circuit. After a short while, the left channel started whistling, a puff of smoke appeared, and the TAD rectifier tube (or so it seemed) stopped glowing. I immediately turned the unit off. The TAD tube is likely ruined. The JJ filter capacitor is rated at 60 µF / 500 V.

I will get around to taking measurements later; unfortunately, I do not have an oscilloscope available.
 
How long was the unit off before turning it on it's side and back on? Tube amp power supplies don't like being turned off and back on within a minute or so. Causes surges with power supply capacitors. Even if the tubes were unplugged, the power supply was working. Could have caused the 5AR4 to arc. I have popped fuses doing that. Just a thought. Paul.
 
I let the vacuum tubes cool down so I could pull them out of the Ampat sockets; I estimate I turned the device back on after about 10–15 minutes. I also think there was probably some sort of random occurrence, like a part coming loose...
 
After working on one of my tube amps, it would scream when I turned it on. Knew it was something I did. Caused me fits until I found with a magnifier where a drop of solder from the iron had landed perfectly on the tiny ceramic disc across the feedback resistor. Sometimes things happen. Finding them can be the hard part. Paul.
 
I was poking around in an amp just taking some voltage readings to confirm sanity and something went bonkers. Found a little blob of solder between a terminal strip and the chassis and touching the terminal with the meter was just enough to push the two in contact. Luckily I knew exactly what I was doing when it went goofy so I had a starting place.
 
You must log in or register to reply here.
Back
Top Bottom