Hi everyone.
I acquired a super-rare NIB un-assembled Heathkit AA-21 amplifier that I just finished assembling moments ago. Being that the kit is so old (even though the box was still sealed), I used new electrolytic capacitors and 1% metal film resistors instead of the originals. I didn't want to chance using the 50+ year old electrolytic caps and carbon resistors that were in the box.
Anyway, I haven't plugged it in yet because the pre-power-on resistance checks are...WRONG.
Resistance measurements at all the test points on the power transistors and in the power supply circuit are all high, and sometimes double what Heathkit says it should be. The instructions say +-20% or don't plug it in.
So my question is this...
It makes perfect sense that if the resistance checks were low, that could indicate a partial or full short resulting in a catastrophic situation.
But what if the resistance checks are high? Is it safe to power up with a variac and a dim bulb tester, or am I going to destroy the germanium transistors in the process?
Thank you in advance for any assistance you care to offer!
Gene
I acquired a super-rare NIB un-assembled Heathkit AA-21 amplifier that I just finished assembling moments ago. Being that the kit is so old (even though the box was still sealed), I used new electrolytic capacitors and 1% metal film resistors instead of the originals. I didn't want to chance using the 50+ year old electrolytic caps and carbon resistors that were in the box.
Anyway, I haven't plugged it in yet because the pre-power-on resistance checks are...WRONG.
Resistance measurements at all the test points on the power transistors and in the power supply circuit are all high, and sometimes double what Heathkit says it should be. The instructions say +-20% or don't plug it in.
So my question is this...
It makes perfect sense that if the resistance checks were low, that could indicate a partial or full short resulting in a catastrophic situation.
But what if the resistance checks are high? Is it safe to power up with a variac and a dim bulb tester, or am I going to destroy the germanium transistors in the process?
Thank you in advance for any assistance you care to offer!
Gene
Different ohmmeters will have different voltages across the probes which may or may not bias on transistor/diode junctions.
The difference between your ohmmeter and their reference (probably a VTVM) may account for the discrepancy.
The difference between your ohmmeter and their reference (probably a VTVM) may account for the discrepancy.
({edit} I think Raichu posted while I was caring for my dog...)
What exact type of meter did Heathkit specify for these checks?
Around semiconductors, today's DMMs will read differently from VTVMs or passive VOMs. DMMs use 0.2V, others often used 1.5V. Either can be right, but use the type that Heath used.
An annotated schematic with theirs/your readings marked may be instructive.
What exact type of meter did Heathkit specify for these checks?
Around semiconductors, today's DMMs will read differently from VTVMs or passive VOMs. DMMs use 0.2V, others often used 1.5V. Either can be right, but use the type that Heath used.
An annotated schematic with theirs/your readings marked may be instructive.
Wow. So here's a link to the schematic and the resistance
Great! I have both. I am just building my confidence before I try that. Thanks...
You know, that's quite interesting. Heathkit explicitly says use a VTVM in bold. I used a Fluke 101. The resistance measurements initially start at what would be normal, and gradually climb to double or greater. Here's a link to high res of schematic and the resistance check with my measurements typed in so you can see what I am talking about. The very high resolution files are too big to upload so I used WeTransfer.
https://we.tl/t-XYet3OOnuO
A VTVM might be something I should get anyway as I have many vintage tube amps that I want to work on.
Good point on the file size. I made them huge because the print in the schematic quickly became unreadable.
Thank you for your response!
Good point on the file size. I made them huge because the print in the schematic quickly became unreadable.
Thank you for your response!
I have found modern, post 2010, DVM to be useless on the resistance scale, unless actually measuring resistance of a resistor. All semiconductor junctions are ----- because the last meter I had that measured at 2 v was a 2000 model sears craftsman That was stolen in 2018.
This is getting to be such a problem that I bought some 50 ua analog meters from a surplus house, and am going to power them with 3 v batteries through series resistors, for resistance checks. In particular I can't find which leg of LED is plus, because "diode scale" of klein or tenma DVM from 2018 do not have enough voltage behind it to make any LED conduct. I used to estimate leakage of electrolytic capacitors by time to charge up on 20 mohm scale. This test on modern DVM always give the same answer. ----.
Your link is useless, so I cant tell if heathkit had you testing any semiconductors or large capacitors with your test. If you are testing only resistors, things might match up, although modern DVM at ultra low voltage don't break across the oxide layer of copper/tin/brass very well. Plus the klein meter sold at HD, the original shielded probes are too short to touch any microelectronics and replacement banana plug probes fall out of the sockets all the time. So **** safe as to be useless.
One problem of modern metal film resistors, is that they do not have the tolerance to surge current of carbon comp resistors of the seventies. Several modern substitutions have resulted in repeated burnout even though the modern metal film resistor had the recommended wattage of the 1970's original. When substituting metal film resistors try to buy a resistor that has the same physical length of the carbon comp. This will usually result in a resistor of several times the carbon comp wattage rating.
This is getting to be such a problem that I bought some 50 ua analog meters from a surplus house, and am going to power them with 3 v batteries through series resistors, for resistance checks. In particular I can't find which leg of LED is plus, because "diode scale" of klein or tenma DVM from 2018 do not have enough voltage behind it to make any LED conduct. I used to estimate leakage of electrolytic capacitors by time to charge up on 20 mohm scale. This test on modern DVM always give the same answer. ----.
Your link is useless, so I cant tell if heathkit had you testing any semiconductors or large capacitors with your test. If you are testing only resistors, things might match up, although modern DVM at ultra low voltage don't break across the oxide layer of copper/tin/brass very well. Plus the klein meter sold at HD, the original shielded probes are too short to touch any microelectronics and replacement banana plug probes fall out of the sockets all the time. So **** safe as to be useless.
One problem of modern metal film resistors, is that they do not have the tolerance to surge current of carbon comp resistors of the seventies. Several modern substitutions have resulted in repeated burnout even though the modern metal film resistor had the recommended wattage of the 1970's original. When substituting metal film resistors try to buy a resistor that has the same physical length of the carbon comp. This will usually result in a resistor of several times the carbon comp wattage rating.
Last edited:
Test on post 7 tests transistor junctions you might get similar results from diode scale of a modern DVM, as long as transistors are silicon. I'd say any silicon junction between 400 mv to 750 mv forwards is worth a try. Germanium transistors may give wrong results, modern DVMs are not tested on those.
Definitely put a 60 w bulb series the AC input line when powering this up. Light should flash briefly then stay out on power up. Should be able to get a watt or two out on load resistor before the bulb starts lighting up. I use load resistor on speaker terminals for initial tests, instead of speakers. Resistor doesn't have to be 8 ohms, can be higher, though the 8 ohm 225 watt logs I have are useful for measuring power out versus spec. P=(v^2)/z where z is load impedance. z = resistance on resistors, z=~4/3 resistance on speakers.
V of music for power test has to be tested with analog meter, RMS capable digital meter, or oscilloscope. Analog meter is the cheapest, and won't lie to you if ultrasonic oscillation occurs above 7 khz the way a fluke meter will.
Definitely put a 60 w bulb series the AC input line when powering this up. Light should flash briefly then stay out on power up. Should be able to get a watt or two out on load resistor before the bulb starts lighting up. I use load resistor on speaker terminals for initial tests, instead of speakers. Resistor doesn't have to be 8 ohms, can be higher, though the 8 ohm 225 watt logs I have are useful for measuring power out versus spec. P=(v^2)/z where z is load impedance. z = resistance on resistors, z=~4/3 resistance on speakers.
V of music for power test has to be tested with analog meter, RMS capable digital meter, or oscilloscope. Analog meter is the cheapest, and won't lie to you if ultrasonic oscillation occurs above 7 khz the way a fluke meter will.
Germanium. At least none test shorted, though, so the only tin whiskers so far are mine own. After I go over the wiring, I think I may give the dim bulb a workout using suitable load resistors.