Hi all,
Relatively new to the forum, and VERY novice with some familiarity with electronics.
I picked up an old "Amherst A-2000" 135watt amplifier relatively cheap to use in the gym at a USFS Fire Station.
I understand that this amp and company were a very short-term 'experiment' formed with Steve Keiser from "B&K" back in the early 90s.
They only made a few (under 50?) but they apparently were great....just cost too much to make for the price point they wanted to offer.
Allegedly, Steve Keiser went onto Luminance Audio to help create their KST-150 which looks EXTREMELY similar, but costs $3000 (still made/offered).
https://www.audiocircle.com/index.php?topic=74950.0
The amp works I 'think' just fine, (again, I am relatively very minimally aware of electronics).
Two items I have noticed though is a mechanical hum from the HUGE torodial transformer, and fair amount of heat.
I took a look at the internals, and I do see some heat damage around/near some resistors caused from what I think are bi-polar transistors with heat sinks (which do get really hot) on the board.
The hum from the transformer is somewhat noticeable from up to 4' away with nothing playing.
The amp was put on a variac for awhile prior to my purchase.
I plugged it in and left it for a day, then turned on with nothing attached for a day, then attached to rest of my system.
Again, it works 'fine' from what I can tell.
I'd like to see if I can help deal with the heat issue from the transistors, as well as the hum from the transformer.
I imagine loosening and re-tightening the hold down screw and bolt for the transformer is where I should start with that, and possibly rotating to see if I can get it to diminish.
How can I help alleviate the heat? Remove the little heat sinks around the transistors clean with alcohol and re-apply thermal paste?
Is there an volt meter test I should do to see if there is an issue with the actual components that is causing the heat?
Images attached.
Last image is from someones sale on USAudio Mart-not my amp.
Thank you for your time and any advice or experience!
Relatively new to the forum, and VERY novice with some familiarity with electronics.
I picked up an old "Amherst A-2000" 135watt amplifier relatively cheap to use in the gym at a USFS Fire Station.
I understand that this amp and company were a very short-term 'experiment' formed with Steve Keiser from "B&K" back in the early 90s.
They only made a few (under 50?) but they apparently were great....just cost too much to make for the price point they wanted to offer.
Allegedly, Steve Keiser went onto Luminance Audio to help create their KST-150 which looks EXTREMELY similar, but costs $3000 (still made/offered).
https://www.audiocircle.com/index.php?topic=74950.0
The amp works I 'think' just fine, (again, I am relatively very minimally aware of electronics).
Two items I have noticed though is a mechanical hum from the HUGE torodial transformer, and fair amount of heat.
I took a look at the internals, and I do see some heat damage around/near some resistors caused from what I think are bi-polar transistors with heat sinks (which do get really hot) on the board.
The hum from the transformer is somewhat noticeable from up to 4' away with nothing playing.
The amp was put on a variac for awhile prior to my purchase.
I plugged it in and left it for a day, then turned on with nothing attached for a day, then attached to rest of my system.
Again, it works 'fine' from what I can tell.
I'd like to see if I can help deal with the heat issue from the transistors, as well as the hum from the transformer.
I imagine loosening and re-tightening the hold down screw and bolt for the transformer is where I should start with that, and possibly rotating to see if I can get it to diminish.
How can I help alleviate the heat? Remove the little heat sinks around the transistors clean with alcohol and re-apply thermal paste?
Is there an volt meter test I should do to see if there is an issue with the actual components that is causing the heat?
Images attached.
Last image is from someones sale on USAudio Mart-not my amp.
Thank you for your time and any advice or experience!
The bias has probably drifted so the quiescent current draw is higher than it should be. There is a preset pot on each amp board to adjust this. You need some way to monitor the voltage across the emitter resistors (but its not obvious which those are without a schematic, one or both of the large ceramic resistors seems most likely). And you need to know the correct bias. Best to adjust the bias with no speakers connected in case of oscillation. Adjust carefully observing the current - making the overbias worse can pull a huge current risking blowing the fuses or output transistors.
I am assuming the output transistors are BJTs or darlingtons - do they have part numbers on them?
The noise from the transformer might well be exacerbated by higher current draw, or perhaps the rubber pad under it has gone hard with age?
I am assuming the output transistors are BJTs or darlingtons - do they have part numbers on them?
The noise from the transformer might well be exacerbated by higher current draw, or perhaps the rubber pad under it has gone hard with age?
the transistors appear to be K**58 and J***2. If these are the legendary 2SK1058, 2SJ162, they are legendary parts you can't buy anymore. Seems as if half the diy designs on here specify those, despite being unobtanium for 20 years.
Definitely get the idle current right you don't want to blow these up.
If idle current is approximately 20 ma, which would be safe on those packages, I would add a fan or a heat sink with more surface area, taller or with more fins. Renew insulators if you lift the transistors, 30 year old mice can fracture, especially if moved.
Since the heatsinks are outside the case, you could easily add taller ones. 4 holes for transistors have to be accurate; the mounting holes less so.
Definitely get the idle current right you don't want to blow these up.
If idle current is approximately 20 ma, which would be safe on those packages, I would add a fan or a heat sink with more surface area, taller or with more fins. Renew insulators if you lift the transistors, 30 year old mice can fracture, especially if moved.
Since the heatsinks are outside the case, you could easily add taller ones. 4 holes for transistors have to be accurate; the mounting holes less so.
Last edited:
Ooooffff. Ok. I do have a Fluke DVM.
I see what looks like a trim pot on each board, and I assume this is what you are referring too in 'adjusting the bias'?
I have no idea how one could even find what the proper bias was/is suppose to be?
Do I not need a 'load' for the amp to see in order to properly measure the bias?
The output transistors do have markings, and somehow I totally forgot to photograph those, so I will pop open and capture.
According to this site: ""Top Quality Hitachi Complementary Lateral Mosfet Output Stages, 4 L-MOSFETS per channel. High Quality Gold Input & Output jacks. All steel chassis with Alum anodized heat sinks DC coupled Circuitry with DC servo offset nulling. Adjustable output stage DC bias (300 mA factory setting). Adjustable input offset stage for ouput DC offset nulling. Input power spec. = 115 VAC, 50-60 Hz only."
There seems to be a thin rubber mat under the transformer, but I really do wonder if there is an existing electrical issue causing it to hum?
@indianajo I'll try and write down and photograph closer each one of those that you mentioned. Chances are they are some form of unobtanium!!!
What is the best way to properly measure the idling current?
What is the best way to properly measure the idling current?
Short the inputs or leave them unconnected. Turn down the volume. Disconnect any speaker. Between the drain of each pair of output transistors, there should be a couple of medium wattage sub-one-ohm resistors. Drain is the leg on the right of the transistor with the lettering facing you and pointed up. The middle of the two drain resistors is the speaker connection. You take a couple of alligator clip leads, measure across one of those resistors, voltage. V/R=I where I is current. .02 A or 20 ma is usually a safe idle current on TO-247 or TO-3p packages (about 7/8" wide). I'm not an expert on lateral fets; somebody that has an amherst manual or another model with j162/k1058 may chme in with a different value current. Since there is no speaker, you may assume the current in one drain resistor is the same as the opposite one. K1058 is paired off with J162. You adjust the pot to get the voltage equal to 20 ma. Do not short across the leads of the transistor with your probe or clip lead. I use Pamona grabbers myself, to limit the amount of damage I do.
The other channel will have it's own idle current, and control pot.
You blurb post 4 mentions 300 ma idle current. Don't know. It mentions DC offset pot, which IMHO should be adjusted to <0.2 v on speaker, lower if you can get it.
There are lateral fets available, but from another manufacturer and in USA, only available in TO3 package. Drilling a heat sink accurately for TO3 is a stretch of my drill control skills. Also TO3 the leads go on the back of the mounting surface, which is the fin side of your heat sinks. Not really a possible conversion for a newbie IMHO. There is the mythical lateral fet Alfet alf16v alf20v in TO-247 package, but I've never seen any in stock at a distributor I deal with. Some people import them from Europe I hear, which involves customs.
Some people say DC on a AC toroid transformer makes it hum. If you are out in the country in Canada, this is a possibility. There are isolation transformers to eliminate the DC, but itself may hum. There are whole threads about this under power supplies topic. My fleet are all old e-core transformers, the heavy obsolete (for bar bands that set up every night) kind.
The other channel will have it's own idle current, and control pot.
You blurb post 4 mentions 300 ma idle current. Don't know. It mentions DC offset pot, which IMHO should be adjusted to <0.2 v on speaker, lower if you can get it.
There are lateral fets available, but from another manufacturer and in USA, only available in TO3 package. Drilling a heat sink accurately for TO3 is a stretch of my drill control skills. Also TO3 the leads go on the back of the mounting surface, which is the fin side of your heat sinks. Not really a possible conversion for a newbie IMHO. There is the mythical lateral fet Alfet alf16v alf20v in TO-247 package, but I've never seen any in stock at a distributor I deal with. Some people import them from Europe I hear, which involves customs.
Some people say DC on a AC toroid transformer makes it hum. If you are out in the country in Canada, this is a possibility. There are isolation transformers to eliminate the DC, but itself may hum. There are whole threads about this under power supplies topic. My fleet are all old e-core transformers, the heavy obsolete (for bar bands that set up every night) kind.
Last edited:
Thanks a ton for all the great info @indianajo !!!
Though I'll be honest, I think I know and grasp maybe 1/10 of what you mentioned!! 😀
The output transistors are the black 'tombstone' three leg devices attached to the back of the amp, yes?
I saw there are two pots I believe on each board (in the photo they are bent vertically, not laying flat (black and white).
Though I'll be honest, I think I know and grasp maybe 1/10 of what you mentioned!! 😀
The output transistors are the black 'tombstone' three leg devices attached to the back of the amp, yes?
I saw there are two pots I believe on each board (in the photo they are bent vertically, not laying flat (black and white).
Yes. The drains are on the right leg. I think the drain resistors may be those longer ones near the O.T.'s that are maybe violet, brown, gold? That would be 0.71 ohms? Maybe violet green gold, 0.75 ohms. One end of each drain resistor should be zero ohms to the "out" terminal for the speaker. Your diode scale on the DVM should beep when you find zero ohms. The other end of each drain resistor should be zero ohms (beep) to the drain of the transistor.
Basic safety for newbies: Don't use 2 hands to check things with the DVM. Use an alligator clip lead on one probe. >24 v across your heart can stop it. No jewelry on hands, wrists, or neck. 1 v at high current through a ring can burn your flesh to charcoal. Wear safety glasses, parts explode sometimes. Also solder splashes when unsoldering particularly.
unplug the unit or turn off with a power strip before touching metal. Measure at <1 v to speaker ground before touching metal. Make sure a power strip, the big blade is actually the neutral (0 volts to the safety pin). Many rural houses have a fake safety pin that isn't connected to anything. Many rural houses (I had one) have the big blade, the neutral, actually hot. Don't work on electronics with one of those. Correct the house wiring before fixing the electronics. (The chassis of the TV I was working on was hot).
Last edited:
Many thanks @indianajo !!!
Greatly appreciate the advice and tutorial!
I was able to get a copy of the schematic thru a contact who got it years ago from Steve- but sadly the file is so small when I zoom in to read values I can't due to resolution.
But I will try to read and decipher what I can, and follow your instructions.
I think I follow a couple youtube electricians who work on older amps a lot, and see if they show basic amp testing just to gain more familiarity and awareness.
First step it clip-leads for meter!!!
Greatly appreciate the advice and tutorial!
I was able to get a copy of the schematic thru a contact who got it years ago from Steve- but sadly the file is so small when I zoom in to read values I can't due to resolution.
But I will try to read and decipher what I can, and follow your instructions.
I think I follow a couple youtube electricians who work on older amps a lot, and see if they show basic amp testing just to gain more familiarity and awareness.
First step it clip-leads for meter!!!
Okay, buying tip for Canadians. Don't buy a shipment from US by either UPS or Fed-Ex ground. The brokerage fee for customs is ~$40. USPS+RoyalMail only. Digikey in MN has a truck they run over the border daily, they handle customs. Parts-express had a nice 10 pack of 60 v alligator clip leads for ~$10, also a useful vari-temp soldering iron for ~$40 but they are addicted to UPS surface I'm afraid. Newark allows you to choose USPS priority in the shipping choice. Farnell (Newark worldwide) has an office in CA but I think they ship from US. There are small electronics houses in Vancouver, Bob's or something like that. Somebody on parts forum recently said UPS air handles the customs without the $40 fee.
To save shipping charges, you might buy replacements for those tiny capacitors with the red ends now. Say 33 on them, that might be a part #. All the tantalum ones like that in my 1970s organs were bad. Maybe the manufacturer got better by the nineties, I don't know. All the 80-s 90s tantalums I have seen were shaped like tan M&M peanut only smaller.
You've got some 30 year old electrolytics that are doomed to fail, too. Rubber decays working or sitting on the shelf. Those 2 uf caps are MKP, forever caps. Electrolytics have + on one end, minus in balls near one lead, or NP after the working voltage. I like >3000 hours service life electrolytics. when I was buying at the TV parts store I had to replace electrolytics every 7-8 years. Digikey will show you the hours service life in the selector table, Newark is now limiting that to the description.
When changing electrolytics, mark + on the board with a sharpie before removal. If you put them in backwards they blow the top. NP electrolytics are non-polar, either way is fine.
To save shipping charges, you might buy replacements for those tiny capacitors with the red ends now. Say 33 on them, that might be a part #. All the tantalum ones like that in my 1970s organs were bad. Maybe the manufacturer got better by the nineties, I don't know. All the 80-s 90s tantalums I have seen were shaped like tan M&M peanut only smaller.
You've got some 30 year old electrolytics that are doomed to fail, too. Rubber decays working or sitting on the shelf. Those 2 uf caps are MKP, forever caps. Electrolytics have + on one end, minus in balls near one lead, or NP after the working voltage. I like >3000 hours service life electrolytics. when I was buying at the TV parts store I had to replace electrolytics every 7-8 years. Digikey will show you the hours service life in the selector table, Newark is now limiting that to the description.
When changing electrolytics, mark + on the board with a sharpie before removal. If you put them in backwards they blow the top. NP electrolytics are non-polar, either way is fine.
Last edited:
Hi Mark and @indianajo
@indianjo I followed some of your instructions, and this is what I was able to note so far.
The transistors on the back heat sinks are (left to right looking at front of amp) K1058 2c1, K1058 2C1, J162 2E1, J162 2E1. Same repeated on next board. (you were correct on those) Hitachi I think if I remember right?
There are 2 pots on each board: assuming one if for bias and the other for off-set?
The 4 brown "heated" looking resistors I measured with the amp on: they measure around 67.3v (dc).
Two measure positive and two measure negative, exact same amounts.
The larger blue resistor right in front of the P2 "Pot", (inbetween the two larger horizontal black capacitors) measures 132v (same on both boards).
I can measure that pot = 1.44v
The smaller blue resistor infront of the other "pot" measures .55v
That pot measured .072v (measuring across the legs that I can clip to on the pot itself)
I noticed that I could spin the transformer a little, and doing so changed the "hum/buzz" considerably, and then I tightened the retaining bolt.
The hum/buzz is still there, but it's very slight. You'll notice it in a quiet room when you first turn the amp on because you're close to it.
Good about the hum improvement.
All the resistors you measured appear to be thousands of ohms each and not useful for measuring output transistor idle bias current.
The four blue resistors near the output transistors all appear to have one leg in common. That probably leads over to the 1 ohm power resistor and the RF block coil before the speaker output. I can't see the colors of those resistors very well, but the 3rd one from either end might be gold. ie 10-2 multiplier. So if violet green gold, could be 75*.01 = 0.75 ohms. A sub one ohm resistor. That is a likely value for a pair of drain resistors, if a little higher than normal.
All the resistors you measured appear to be thousands of ohms each and not useful for measuring output transistor idle bias current.
The four blue resistors near the output transistors all appear to have one leg in common. That probably leads over to the 1 ohm power resistor and the RF block coil before the speaker output. I can't see the colors of those resistors very well, but the 3rd one from either end might be gold. ie 10-2 multiplier. So if violet green gold, could be 75*.01 = 0.75 ohms. A sub one ohm resistor. That is a likely value for a pair of drain resistors, if a little higher than normal.
Okay, use amtprod private messaged me a fuzzy copy of the schematic. The drains of the output transistors are connected together and directly to the input of the RF block coil and 1 ohm resistor (to speaker). There are no resistors on the sources of the output transistors, either. So one cannot measure voltage across the drain resistors to measure idle bias current.
The only way I see to set the output transistor idle bias current is, with the AC power off, and rail caps 0 volts, to pull F1 (or F2) . Using alligator clip leads, connect the amps scale of the dvm across the fuse clips. Use 2 amps scale. Turn the AC power back on again. If the blurb in post # 4 is correct, set the idle current (across the fuse clips) to 300 ma.
Turn AC power off again, check that the rail caps are discharged, then disconnect ammeter and put the fuse back.
Repeat for the other side. Pictures of board are for only one channel, with 2 SK1058 and 2 SJ162 transistors.
The only way I see to set the output transistor idle bias current is, with the AC power off, and rail caps 0 volts, to pull F1 (or F2) . Using alligator clip leads, connect the amps scale of the dvm across the fuse clips. Use 2 amps scale. Turn the AC power back on again. If the blurb in post # 4 is correct, set the idle current (across the fuse clips) to 300 ma.
Turn AC power off again, check that the rail caps are discharged, then disconnect ammeter and put the fuse back.
Repeat for the other side. Pictures of board are for only one channel, with 2 SK1058 and 2 SJ162 transistors.
THanks @indianajo for taking a look. I think I know and understand the process you're mentioning and will try out once I get clip leads for my meter. So measuring across the fuse is closer to what is being driven to the output it sounds like, since there are no specific drain resistors for each output transistor (like in an Adcom 555 as example).
It is the only way to measure idle bias current across the output transistors, as I read the schematic. As before, volume turned down, speaker disconnected, input shorted if you can do that. Any current flowing in through F1 will be passing through the 4 output transistors and also be flowing out F2.
300 ma across 4 pairs of O.T.'s makes more sense than 300 ma across one pair. 75 ma each doesn't sound so scary.
300 ma across 4 pairs of O.T.'s makes more sense than 300 ma across one pair. 75 ma each doesn't sound so scary.
Last edited:
F1 and F2 would be the fuses (one each board) right near the "Vout VPOS" on the schematic, yes? SO hard reading that dang thing!!
In the picture post 1 F1 obviously connects to V+ input. F2 obviously connects to V- input. Removing either gives you a place to put your ammeter. The input & driver stages run off these too, but through R12 R13 R16 & R17, As these are hundreds or thousands of ohms, the amount of current flowing that way will be 1/20 or less of that flowing through the eight output transistors (at idle).
If you download the schematic in *.bmp file, open it with viewer, then your file viewer should be able to expand them by clicking plus or something. Is fuzzy but almost readable.
If you download the schematic in *.bmp file, open it with viewer, then your file viewer should be able to expand them by clicking plus or something. Is fuzzy but almost readable.
For some reason on my computer (mac) even trying Photoshop when I zoom in the schematic is fairly illegible...then again I am barely aware of what I am looking at. But I'm cobbling it together.
So I was able to to the DC reading at the speaker outputs at least, and got .022v right channel, .004v in the left channel. No input signal, no load at all.