Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Sunn Amplifier Power Amp Puzzle

Status
Not open for further replies.
Just a thought Bud,
Did the original transistor have an mica insulator and did you recuse it and use thermal grease with the replacement trans. ?

Added: What trans are you subbing with and what was th eoriginal number?
 
Someone had said the big caps usually don't go bad, and I didn't have any those size, but in looking at them now they are only rated 50V. Not much headroom on a +/- 45 supply. Plus I'm assuming you guys agree that 2vp-p ripple when the supply's not even all the way up is too much, yes?

I think 2 V p-p ripple is too much and I know the "hard way" that 50 V is too little. At 33 years your pushing it.

I'm assuming that the big caps are can type. Mouser has some: https://www.mouser.com/Passive-Comp...0wlgmZ1z0wla7Z1z0wqfaZ1z0wqsuZ1z0x7xmZ1yx4atu

You can always increase the capacitance.

Way back int he 1980's I built my own version of "The Leach Amp" https://leachlegacy.ece.gatech.edu/lowtim/ and I accidentally made a "mirror image" of the PCB. Finding that mistake was really tough. In the end, I was able to use the boards with few changes. My amp has 40,000 uF of filtering and independent power supplies with a custom transformer. The first version had the same filtering, but with an 18 A +-50 V constant voltage transformer

I had a stint (side job) of commercially repairing amps at home.

The stuff I repaired sometimes at work would make your head sizzle. You know the good ole CRT display, When you ramp the power up to 30 kW, the electron beam can melt metals. The unit plugs into a 60 A 208 3 phase outlet and is quite lethal. The high voltage was 15 kV at 1.5 A, not 40 kv at a 10's of milliamps.

amp_outside_front.jpg amp_inside.jpg
 
Just a thought Bud,
Did the original transistor have an mica insulator and did you recuse it and use thermal grease with the replacement trans. ?
Added: What trans are you subbing with and what was th eoriginal number?

Hi Pfofit! The originals had mica insulators with thermal grease. The replacements came with silicone pads. I decided to use those temporarily since I'm still in troubleshooting mode and I'm not pushing the amp at all as I know it has several unseen issues.

The originals were SJ954's opposite 2n3055's. Now I'm still using the 3055's (new ones) but the SJ954's I subbed with MJ2955's. When the amp is done, I'm going to put better brands in, but for testing I picked up 6 each of them from China. There are 3 power transistors on each side, so I know they do work. I've already blown two of the 2955's but it was my own fault.

Thanks for joining in. The more input the better ;-)
 
I'm assuming that the big caps are can type.

Actually, I had said "can" just speaking loosely, but they are axials. I re-ran your search changing to "axial" and only 2 results, neither in stock, and neither with a minimum of < 1000. I appreciate your checking there for me, in any case.

You can always increase the capacitance.

I think my Digi-Key search was for 3000-4000uF. It's amazing how little comes up for axials in that cap range. You still think 63v is pushing it?

Way back int he 1980's I built my own version of "The Leach Amp"

I'm envious! That is a long-range goal of mine . . . after I get good at not blowing things up. Nice work! I think my first homemade PCB was a mirror, too ;-)
 
Yes 63 is pushing it, but it beats 50. 105 C, if you can, Low ESR if you can. I may have some other places up my sleave.

Arrow has some radial caps. https://www.arrow.com/en/products/lgy1k392melb40/nichicon No size, so I can;t narrow. You can make them bigger in value.

You might be able to secure with ty-wraps and holes or a ty-wrap holder. if you have room, I'd even consider the cans. the cans have clamps you can buy.
 
Well, my/the Leach had two 100 uF 50 V capacitors on the amp PCB and my first build had a ferroresonent transformer which turned out to be too noisy, but the bass sounded great. By the original design, (3A)fuses on each supply rail, one (3A) on AC power in and two (3 A) AGX fast blow fuses for speaker protection.
The original article was in Audio Magazine sometime around 1976.

It was built open-frame w/o covers, and I won't do that again. One day it was humming along and I heard POW! The 50 V rated electrolytic exploded. It should have never been 50.

That's when I decided to make a 2U case from scratch, use dual supplies and deal with the slow turn-on of 40,000 uF of capacitance, I wanted the amp to shut-down if a rail fuse blew. The slow turn-on circuit is quite dumb, but it gets the job done. AC is applied through a flame proof resistor initially with the speakers and inputs disconnected until the voltages across each cap is ~2/3 of 50 V. When that happens, the resistor is shorted out, the speakers are connected and the audio ramps up exponentially using an OPTO-FET in series with the input.

The caps are 4 x 9600 uF @ 75 V, 85V surge; computer grade; with a ZNR across each of them. The ZNRs protect my protection circuit. There is an independent 12 V supply for the protection circuit.

If a fuse pops, the resistor will pop. Now, the coolest part was that at one time I reversed the NPN and PNP output transistors and there was no damage but the fusable resistor. That circuit could be a "little smarter". There is no power switch on the amp because it's controlled by a odd timer that was a construction article in Radio-Electronics, It had two modes. A) 0-2 hrs on and B) 0-10 min until off (retriggerable). The B mode had a filter for FM hiss.
So, you had a sleep timer and something that would turn the system off after 10 minutes when the station went off the air, the turntable stopped or the cassette deck stopped.

The boards were commercially made with a ground plane. No silk screen. The art-work was taped.

My upgrades were minor except the power supply and turn-on system. Resistors were nearly all metal film and the bias pot was made 10 turn. Extension cables are needed to work on the amps outside the case.

They specified a 70 V CT at 3A transformer for stereo and I made each channel that. The transformer could use a higher rating. The article suggested electrostatic speakers where the supply would be fine. I eventually found a 500 W AC sine wave regulator for $100 in NYC Canal Street and it's been good.

Regards.
 
Last edited:
Well, my/the Leach had two 100 uF 50 V capacitors on the amp PCB and my first build had a ferroresonent transformer which turned out to be too noisy, but the bass sounded great.

What was its main function for you? PA system? Hi-Fi amp? Just checked out the link and the main circuit diagram. Cool that people are keeping it going after 40 years.
 
Even if there's something wrong with it, it's probably a simple fix.

Well, got lucky with this one, I think. The thing was truly brand new. Instructions and test leads still in sealed baggies and all. Ran it for a while this afternoon, playing with freq, amplitude, the different wave forms. Looks like its good to go when I need it. Which I will when I get back to the preamp board. I know there's one channel that was giving me a hard time about passing a signal. I'd replaced an op-amp and it started working again. Then kaput again. But I'm not quite back at that bridge yet . . .
 
What was its main function for you? PA system? Hi-Fi amp? Just checked out the link and the main circuit diagram. Cool that people are keeping it going after 40 years.

As a 10 YO, I had a Fisher model 50-A tube amp with a heathkit pre-amp. It sounded nice until i drilled a hole through the winding of the intersrage transformer, so I scrapped it. I learned some tube repair skills. It is a sought after tuner, even now. ebay had one for about $700 USD recently. A modified cloone goes for about $10.000 each new. So, I guess I knew what quality sound was. We used to have tube radios around the house. They need some serious work.

When I was putting audio together, I could not afford much, My parent bought an AM/FM Electrophonic Stereo with detachable speakers and a turntable with a ceramic cartridge. Grand total, maybe $150.00. It didn't sound good and had poor FM reception. New, won speakers made the amp sound worse.

I don;t remember the order, but I found a nice power/preamp at a garage sale and I gut the power amp section and I built an RE amplifier that was supposed to be for car based on the NE-540 chip. I had to wind the transformer for that one.

I took a rather slow upgrade path. The first component was a Technics Sl-1700 direct drive semi-automatic turntable. It ejected and parked the tome-arm at the end of playing. It's like, why buy albums if the player will damage them. Invested in record cleaning kits and finally a moving coil cartridge.

I effectively ended up with 3 pieces from this Technics Professional Series: https://www.preservationsound.com/?p=6908. The FM tuner which is fantastic, the pre-amp and the EQ. The pre-amp has a frequency response from 0 to 100 kHz. With the tuner and the rotator, I could receive a different station on the same frequency by turning the antenna. At some point I added the kit Radio-Electronics Timer as a sleep timer and I built a 12 band EQ that was in RE for about $100.00. It too had a frequency response out to 100 kHz.

To make the FM sound even better I acquired a 4bx dynamic range expander from dbx. Retail was $1000. It expanded in 3 bands and had impact restoration, so you could adjust and hear keys hitting the strings on a piano. And I acquired the carver TX1-11 FM charged-coupled decoder which was an external signal processor for FM multi-path.

Cassettes were also popular, so I got a Technics RSB-100 with direct drive. With metal tape and dbx compression, it could achieve up to 25-28 kHz response. dbx was superior to Dolby but rare. Longevity and no belts were primary.

Unfortunately, two portable CD players got either stolen or lost. One had optical out and an IR remote. I miss both of them. I was given a CD player whose motor sticks and I found a changer one for free without a remote.

One of those plans that never materialized was to record FM onto super VHS tapes. I bought a JVC HRS-7000U super VHS recorder with a switched outlet
that could record audio only.

Things now just aren't the same anymore.




I acquired the pre-amp new and the EQ used.


https://www.preservationsound.com/?p=6908
 

I ordered these and some low-ESR 10uF ones, too.

Haven't been able to work on amp. Cat and fridge broke at the same time. Cat's been to vet's. I've got parts on the way for the fridge. Should be able to revive this thread again soon!

Thank you everybody for all the help. I'll post back after I put in the new caps. Coincidentally, I'm pretty sure it's the start cap on my fridge that's also bad. It's the month of bad caps I guess . . .
 
Thoughts:
Replace C2, C3. C7, C8 and C13
Replace C16 and C17 and upgrade voltages to at least 80.

Well, it's been a somewhat rough couple weeks. My cat's still sick and will likely be leaving us soon. But today I found time to get those big caps in. I'd bought low ESR 3900uF 80V caps. Sunn was cool enough to have holes and solder pads already there for them, so that was a nice surprise.

Replacing C16 and C17 got rid of the 1.5v and 2v ripple. Now the ripple at C16 and C17 measures in the mV range ~200mVp-p on one and ~80mVp-p on the other, which I'm guessing is just fine.

At 75VAC at the Variac, the +/- 15vdc supplies read somewhat close to each other @ +14 and -13. I kept cranking the AC till I got the plus side up to +14.9 with 90VAC (didn't seem to want to go a full +15 so I left it there). However, the minus side still read -13vdc.

I'm not 100% sure I changed C13 so I'll change that soon just to rule it out.

I figured out I could plug my Kill-A-Watt into my variac to measure current (since you had said a current meter at the variac would be a good thing). Amp drew about 110mA in the foregoing situation. When I cranked things up to 90VAC it drew 130mA or so. At this point the 40W bulb in the limiter did get pretty bright for 1/2 second at power-on then continued to glow very faintly (barely perceptible in daylight). I don't like the sudden flash, but I guess that's normal till IC1 kicks in and balances the output transistors?

I don't know why that negative supply doesn't want to move up to -15. The rails read dead-balanced at all voltages tried so far. I don't know what's causing the imbalance in the regulated +/- 15vdc supplies. I've already changed the zener and TIP30 to no avail.

I'll post later as to whether the numbers improve after I change c13 . . .
 
Check voltage across CR5, Check R25. Check the voltage across C13. The -15 V ripple.
Compare the voltage drop across R23 to R3. For now.

Things don't look too bad.

I'd really like you to place a short across C13 and get the same numbers you got before. e.g. the +-15 V supply values and the line current at 90 VAC The current measured should go down.
 
At this point the 40W bulb in the limiter did get pretty bright for 1/2 second at power-on then continued to glow very faintly (barely perceptible in daylight). I don't like the sudden flash, but I guess that's normal till IC1 kicks in and balances the output transistors?
It could also be the current surge to charge all the large power supply filter capacitors.
 
It could also be the current surge to charge all the large power supply filter capacitors.

Thank you, Crutschow,

As long as nobody's worried about it, that makes me feel better. Hadn't thought about the caps charging. That's a good point.
 
The old rules:

The voltage across a capacitor can't change instantaneously.
Corollary: It's the ESR of the capacitor which is in parallel that the rail sees first.

The current in an inductor can't change instantaneously.
 
Check voltage across CR5, Check R25. Check the voltage across C13. The -15 V ripple.
Compare the voltage drop across R23 to R3. For now.

Figured I'd do that before changing C13 for a before-and-after. With the Variac @ 90 VAC:

CR5 Vdrop: 13.58vdc
R25 Vdrop: 4.22vdc (resistor measures 132 ohms, which is right at the upper end of its 10% tolerance)
C13 Vdrop: 3.32vdc (C-E of Q13)
R23 Vdrop: 17.2vdc (.1 to .2 a bit jumpy)
R1 Vdrop: 15.05vdc to 15.1vdc (a bit jumpy)
R3 Vdrop: 12.44vdc (Turns on at 11.8vdc. Then, after a 3 sec pause, rises to 12.44vdc over the next 4-5 secs)

Couldn't really bring up a measurable ripple on the -15vdc supply. I will say that it looks "noisy" but I'm using a new digital 'scope, which I'm not used to, so I don't really know what it's supposed to look like. The "measure" function says there's around 20mVp-p, but very jumpy, noisy, doesn't look like a sine or sawtooth, doesn't dial in at 120 hz either.

Interesting that the drops across R23 and R1 vary by about 2.1vdc, which is the same difference by which the regulated supplies differ, just about (+14.9 & - 12.9 unbalanced by 2vdc).

After I found C13, I do think it's one of the few (if not the only) electrolytics left unchanged. I'll do that soon . . .

Thanks for all the input!
Bud
 
One more voltage at 90 VAC: Voltage across R21 and R50;

R21 Vdrop: .2mV
R50 Vdrop: .lmV

Not sure how accurate my meter is down that low, but with the power off they each read 0.0 and these are the readings I get with 90VAC. That's the last decimal place on my meter, so it could mean .14 & .16 with traditional rounding, but I doubt it's that accurate in any regard. I suspect as long as these voltages aren't very high then things are kosher?
 
Thanks. The amps has no idle current yet and I think it should by now. R21/R50 should have 90 mV per the schematic. See the 90 mV note in the bottom right corner.

Right now, I think that your -15 V supply is being overloaded because of another problem.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top