Sanui 3000A Volume Control and Output
- Thread starter rickygallagh
- Start date
kevzep
Its all about the Music
Are you using two meters so you can look at both negative and positive rails at the same time?
Also leave the fuses out of the other channel just in case the bias is running away on those without you knowing.
But this is what these amplifiers do, it will have you running around in circles......it will consume a lot of your time getting it right and then it will throw a curve ball just when you think it is all going well.
Also leave the fuses out of the other channel just in case the bias is running away on those without you knowing.
But this is what these amplifiers do, it will have you running around in circles......it will consume a lot of your time getting it right and then it will throw a curve ball just when you think it is all going well.
McGowdog
Active Member
What I've done is blown fuses.
So if I hook one meter up to the outer R side, another meter to the outer L side, leave the two inner fuses out, I can try and tweak those pots to get 80 mA on each?
I could do or try this.
Have some 630 mA fuses on order for my two fluke 77s. One of them had an extra built into the case, so I'm down to one measly fuse.
I've tried hard to start off at 10 amp setting before I go to mA range. Can't believe how touchy this bias is.
On my Dynaco Mark IIIs, I tweak a pot until I get 1.56 volts and done.
So if I hook one meter up to the outer R side, another meter to the outer L side, leave the two inner fuses out, I can try and tweak those pots to get 80 mA on each?
I could do or try this.
Have some 630 mA fuses on order for my two fluke 77s. One of them had an extra built into the case, so I'm down to one measly fuse.
I've tried hard to start off at 10 amp setting before I go to mA range. Can't believe how touchy this bias is.
On my Dynaco Mark IIIs, I tweak a pot until I get 1.56 volts and done.
McGowdog
Active Member
Oh wait, you want me to look at + and - rail on one channel then do next rail separately?
This I can do.
I thought I'd seen one of you guys suggesting I look at all four at the same time? If so, how is this done without blowing fuses? Start at 10 amp and work my way down to 250 mA?
Why not a current meter with better resolution? Why can't they offer me 2 or 5 a full-scale?
This I can do.
I thought I'd seen one of you guys suggesting I look at all four at the same time? If so, how is this done without blowing fuses? Start at 10 amp and work my way down to 250 mA?
Why not a current meter with better resolution? Why can't they offer me 2 or 5 a full-scale?
McGowdog
Active Member
F that pos 3000a.
Came back with fuses and got two current meters set up. Turned it on, hmmm, what happened to the lamps?
No power?
Measured dc rails that had pulled fuses. Only about +- 26v. Rear fuse blown. Oh, now an output transistor is finally blown. Ce short, bc and be low, less than 20 ohms.
Checked others, not shorted yet, replaced fuse and checked out rails. Now I have a healthy -+44 again.
Looked for bias current again just because I was set up, about 110 and -110 v.
Can't do much until I get new set of output transistors.
Want to have a good set of four, maybe get a 5th one to complete this test.
Sick of the 3000 for now.
Went and bought a minty g5000 just to take my mind off this one for now.
Correction, didn't replace transistor, just took it out.
2nd correction... +- 110 mA. Was late and tired.
3rd edition...
Oh, I need to be monitoring that darned dc voltage as well, but have been going ol on outputs for now anyway until I get amp healthy enough to do so?
Came back with fuses and got two current meters set up. Turned it on, hmmm, what happened to the lamps?
No power?
Measured dc rails that had pulled fuses. Only about +- 26v. Rear fuse blown. Oh, now an output transistor is finally blown. Ce short, bc and be low, less than 20 ohms.
Checked others, not shorted yet, replaced fuse and checked out rails. Now I have a healthy -+44 again.
Looked for bias current again just because I was set up, about 110 and -110 v.
Can't do much until I get new set of output transistors.
Want to have a good set of four, maybe get a 5th one to complete this test.
Sick of the 3000 for now.
Went and bought a minty g5000 just to take my mind off this one for now.
Correction, didn't replace transistor, just took it out.
2nd correction... +- 110 mA. Was late and tired.
3rd edition...
Oh, I need to be monitoring that darned dc voltage as well, but have been going ol on outputs for now anyway until I get amp healthy enough to do so?
Last edited:
McGowdog
Active Member
Well today is the 10th of June and I finally got some transistors.
Guess I'll slowly and reluctantly investigate what mods have been done on this unit and wbat has yet to be done.
I'm gonna try to at least get this bias done and maybe see what my dc offset is, to 10 ohm 1 watt?
Guess I'll slowly and reluctantly investigate what mods have been done on this unit and wbat has yet to be done.
I'm gonna try to at least get this bias done and maybe see what my dc offset is, to 10 ohm 1 watt?
kevzep
Its all about the Music
If you wanted to protect the output transistors from going Fzzzzzzzzzzzzzzip....poof!! Install some 10watt 100Ω resistors in place of the emitter resistors. They will act as current limiters and save your ass when things go wrong, replace the bias trimmers as well if you already haven't.
I think Sansui want a load on the output to measure the DC offset, hence that 10Ω 1 Watt nonsense...
I think Sansui want a load on the output to measure the DC offset, hence that 10Ω 1 Watt nonsense...
McGowdog
Active Member
The pots are 200 ohms but in parallel with 15 ohm resistors. What's the point in that?
You serious? Change 0.3 ohms out with 100 ohm 10 watt?
kevzep
Its all about the Music
Only need the resistor on the output if its got the mod because that makes it a capacitor coupled output. So no need to put a load on the output to assess DC offset.
Yes I am dead serious about the 100Ω resistors, they will act as current limiters and protect your output transistors and other associated components if something should "go west" on you.
Once the amplifier is stable then you put the 0.33Ω resistors back in.
Yes I am dead serious about the 100Ω resistors, they will act as current limiters and protect your output transistors and other associated components if something should "go west" on you.
Once the amplifier is stable then you put the 0.33Ω resistors back in.
McGowdog
Active Member
Thanks. Is this part of a modification I can obtain documentation and parts lists on?
So far, the only modification I have documents of is service bulletin ref. CE-007-A.
This bulletin has two non-polar 1000uF caps put in series to white and blue wires which are in my understanding on their way to load switch and ultimately the power transistors.
Then, they have you replace 0.3 ohm resistors r16 17 18 & 19 with 1 ohm 3 watters. I believe these are off the emitter of pwr transistors.
Then you add a 10 ohm 1/2 watter to TR 2 3 4 & 5, to their base.
Then you replace Q.A. F1 f2 f3 & f4 from 4a to 2a.
Then you replace diodes d801 through d808 from sm-150 to S-1.5-02.
Then you put in a 10 ohm 1W resistor to each speaker and adjust for the 80 mA.
Maybe they have a typo there. Maybe they meant 100 watt instead of 1.
I have available two 8 ohm loads at 250 watt.
So far, the only modification I have documents of is service bulletin ref. CE-007-A.
This bulletin has two non-polar 1000uF caps put in series to white and blue wires which are in my understanding on their way to load switch and ultimately the power transistors.
Then, they have you replace 0.3 ohm resistors r16 17 18 & 19 with 1 ohm 3 watters. I believe these are off the emitter of pwr transistors.
Then you add a 10 ohm 1/2 watter to TR 2 3 4 & 5, to their base.
Then you replace Q.A. F1 f2 f3 & f4 from 4a to 2a.
Then you replace diodes d801 through d808 from sm-150 to S-1.5-02.
Then you put in a 10 ohm 1W resistor to each speaker and adjust for the 80 mA.
Maybe they have a typo there. Maybe they meant 100 watt instead of 1.
I have available two 8 ohm loads at 250 watt.
Last edited:
kevzep
Its all about the Music
That sounds like the factory mod to make the amplifier cap coupled.
They want you to put the 10Ω 1W resistor on the amplifier output to put a load on it to set the bias. There's no need to have a 100W resistor as the amplifier isn't swinging any voltage its only idling.
Do exactly what they say, don't start getting creative and reading things into the instructions, otherwise you'll have a pile of molten amplifier to deal with...
I know little of the service bulletins and whatnot......the 3000A I had, I gave away with a parts donor as well....It was running when I gave it away, but of course as soon as it was powered up it failed spectacularly.
It did not have the Cap coupled mod...
They want you to put the 10Ω 1W resistor on the amplifier output to put a load on it to set the bias. There's no need to have a 100W resistor as the amplifier isn't swinging any voltage its only idling.
Do exactly what they say, don't start getting creative and reading things into the instructions, otherwise you'll have a pile of molten amplifier to deal with...
I know little of the service bulletins and whatnot......the 3000A I had, I gave away with a parts donor as well....It was running when I gave it away, but of course as soon as it was powered up it failed spectacularly.
It did not have the Cap coupled mod...
McGowdog
Active Member
So what y'all are saying is, None of us can fix this thing, but do exactly as we say.
Got it.
You go from saying 100 ohm 10 watt to agreeing with 10 ohm one watt.
I'm confused.
Having a hard time finding a kit. May have to piece it together myself.
Interesting ideas here. Don't think I'd call it a Sansui 3000a anymore, but it's something.
Got it.
You go from saying 100 ohm 10 watt to agreeing with 10 ohm one watt.
I'm confused.
Having a hard time finding a kit. May have to piece it together myself.
I've made some progress. The new outputs are in and the 1000uf caps for the output mod are installed. I ended up using 3 caps per side, for 3000uf each channel.
I did find what seems a workable method to bias this thing and get DC offset where it should be.
- Remove both fuses from one channel.
- using two meters, connect to measure mA across each of the two fuse holders
- Use a third meter to measure DC offset at a point before the cap mod
- adjust both bias pots for 80mA while watching the 3rd meter for offset
- you'll find a sweet spot where 80mA is achieved on both sides of the one channel while the DC offset lands where it should be.
- Repeat above steps for the second channel
I was able to do both channels this way and have a rock solid 80 mA on both channels and a DC offset of around 1mV each.
It looks like I need a new volume pot, as at full off there is enough volume to hear program material quite well. Measuring the low wiper on the pot shows 180 ohms and 320 ohms (L/R) to ground, where it should be at or closer to 0/ground.
However - it sounds great! A very sweet sound from the little testing I've done so far. I'm very pleased.
Pre-Power board with new diodes, resistors, and multi-turn pots for bias. And below that, the Driver amp board with new caps, pre-driver transistors, etc.
3000uf each channel for the output mod. Still need to find some 25mm fuses...
Ripple-Filter board with new caps.
Phono preamp with new caps. Input cap replaced with a film.
I don't have a picture of it, but I also pulled and recapped the Control amp board and did the caps and transistors there also.
Interesting ideas here. Don't think I'd call it a Sansui 3000a anymore, but it's something.
Last edited:
kevzep
Its all about the Music
You will have to put the kit together yourself, I have never heard of kit being available for these amplifiers nowadays, they were back in the day as Sansui had to cover their ass because of the instability of these amps.
No, I never said that at all!! Read the posts properly my friend, I am trying to help you. I said "If you wanted to protect the output transistors from going Fzzzzzzzzzzzzzzip....poof!! Install some 10watt 100Ω resistors in place of the emitter resistors."
I said this, as any tech will tell to, because they will act as current limiters and save you blowing output transistors during diagnoses.
You said "Then you put in a 10 ohm 1W resistor to each speaker and adjust for the 80 mA."
You also said "Maybe they have a typo there. Maybe they meant 100 watt instead of 1."
Then I said, "They want you to put the 10Ω 1W resistor on the amplifier output to put a load on it to set the bias. There's no need to have a 100W resistor as the amplifier isn't swinging any voltage its only idling."
The 10Ω 1watt resistor has nothing to do with my suggestion of using 100Ω 10watt resistors in place of the emitter resistors!!
The 100Ω resistors will protect the output transistors from detonating on you whilst you are working on it and getting it stable.
Once stable, you can use whatever Sansui suggest to use for emitter resistors.
The one I had, I got it working in stock form, no output mod, it ran for about a month, then it destroyed a pair of Pioneer CS60 speakers.
I repaired it again, and it was working, but then I gave it away.
So, it is possible to get it working!!
McGowdog
Active Member
Thank you for your time. Speaking of time, I'm actually not in a hurry here, just frustrated. I'm willing to take the time to get the right parts together and do this.
I was admittedly confused on the emitter resistor AND output bias resistor instructions.
I'll try both suggestions, while learning their specific purposes.
I don't want to fry speakers down the road, so I will attempt the cap mod at minimum.
I, probably similar to you, don't like to go changing and shotgunning every part on the receiver unless/until there is a need.
I'm hunting down some radial 1000uF bipolar or non-polarized caps now. Not sure of the voltage. At least 50v? Maybe 100v? Don't remember anyone specifying this value.
I was admittedly confused on the emitter resistor AND output bias resistor instructions.
I'll try both suggestions, while learning their specific purposes.
I don't want to fry speakers down the road, so I will attempt the cap mod at minimum.
I, probably similar to you, don't like to go changing and shotgunning every part on the receiver unless/until there is a need.
I'm hunting down some radial 1000uF bipolar or non-polarized caps now. Not sure of the voltage. At least 50v? Maybe 100v? Don't remember anyone specifying this value.
kevzep
Its all about the Music
I would get it working in stock form as best you can, because that is as good as it ever gets with these amps, and then apply the mods.
I have never done the mod, or seen one which has been done, so I am not entirely sure whats involved apart from turning it into a Capacitor Coupled amp.
Is the above information all you have?
Are you sure they want 1000µf bi-polar? will be very difficult to get I am thinking....Polarised caps not a problem and you couple even use some really nice Audio Grade Nichicon Superthrough's....
Probably need 50V output caps for that amplifier...
I have never done the mod, or seen one which has been done, so I am not entirely sure whats involved apart from turning it into a Capacitor Coupled amp.
Is the above information all you have?
Are you sure they want 1000µf bi-polar? will be very difficult to get I am thinking....Polarised caps not a problem and you couple even use some really nice Audio Grade Nichicon Superthrough's....
Probably need 50V output caps for that amplifier...
dly66
Super Member
I have a 3000A as a future project on the shelf I may try to get it working this winter. Its also a 69 model does not look like it has had the output mod. The other project is a similar vintage SX-9000.
I was looking at Mouser and in stock options they have in 1000 uF non-polar highest voltage is 35V. Step down to 470 uF adds 50 and 63 V options. If you want higher voltage/capacitance options may require doing series/parallel combinations. Capacitors work opposite of resistors. Two caps in parallel double the capacitance. I think the capacitance controls the low frequency roll off point.
Their mod may be more like putting a high pass crossover filter ahead of the speakers to block the DC, like a first order tweeter crossover in a speaker. The capacitance increases as the crossover frequency decreases. High DC can burn out the woofer the tweeter will usually be okay.
If you can interpret what they are recommending the guy at select45rpm.com over in England likes this Sansui. He claims ignore published ways of biasing and describes another way to go about it.that could be tried. Work down this page to find the comments on the 3000A:
http://select45rpm.com/pages/hifi/vintage-hifi-reviews.html#9
I have used the recommendations of Conrad Hoffman on this site on setting the bias of my 2000A (1970) and it sounds great and has given me no problems. Of course it has had his amp mods done to it.
I was looking at Mouser and in stock options they have in 1000 uF non-polar highest voltage is 35V. Step down to 470 uF adds 50 and 63 V options. If you want higher voltage/capacitance options may require doing series/parallel combinations. Capacitors work opposite of resistors. Two caps in parallel double the capacitance. I think the capacitance controls the low frequency roll off point.
Their mod may be more like putting a high pass crossover filter ahead of the speakers to block the DC, like a first order tweeter crossover in a speaker. The capacitance increases as the crossover frequency decreases. High DC can burn out the woofer the tweeter will usually be okay.
If you can interpret what they are recommending the guy at select45rpm.com over in England likes this Sansui. He claims ignore published ways of biasing and describes another way to go about it.that could be tried. Work down this page to find the comments on the 3000A:
http://select45rpm.com/pages/hifi/vintage-hifi-reviews.html#9
I have used the recommendations of Conrad Hoffman on this site on setting the bias of my 2000A (1970) and it sounds great and has given me no problems. Of course it has had his amp mods done to it.
McGowdog
Active Member
Interesting. He says something about measuring mvolts across .3 ohm resistors I think, not sure, but what mV value is he looking for?
Says to do 1 and 3 first, or top and bottom, then 2 and 4 or middle ones next.
Here's his quote on that portion....
Biasing is best done ignoring any published way, to use the white resistors on the outputs in the top & read mv across them is best: set all pots midway first, then note the transistors are numbered 1,2,4,3 & adjust the 1&3 first & fine tune on 2&4 then recheck again. Then check DC offset & fine tune, then do the same on the resistors to balance it, takes 2 goes to get right. The amp must have a speaker load connected to read DC offset is within a safe limit after adjusting bias without any load, else it'll be confusing. Use the outer two for DC offset, we got ours steady & under 2mV with a speaker load.
Oh, I've got a shorted npn on the driver board too.
Had the left side biasing just fine, plus side, minus side, back and forth, very small dc offset, then went to right side, saw about hundreds of mA on right side, shut off, and saw I had blown number 3 fuse. Need to replace the sony 2sc756 transistor, at least, parts on order.
dly66
Super Member
Have not gotten to mine yet. My understanding the 3000A is a challenging one. Some seem to say you can get them stable and happy some seem to think they are more trouble than they are worth. Got a 1970 Pioneer SX-9000 not known for being easy to work on either. May go back and forth between the two when the time comes.
I will be getting my Onkyo off the work table tonight I can slip my 3000A on there before the tuner I was going to work on, pop the hood and look at the schematic and translate. I will have to look to confirm but if the current at the fuses they talk about go through those resistors you multiple the current by the resistor value to get a mV value to measure across the resistor for the bias adjustment. Conrad Hoffman gives an alternative to Sansui's manual for adjusting the 2000A, especially if you have done his mods to one, and you measure voltage at the four resistors attached to the outputs. They say do the bias adjustment with no load, which is different than Sansui states. When it comes to the DC offset do that with a load attached. They indicate, and based on past experience probably right, it will take at least two rounds to get both close.
When you deal with old vintage items often you will find bad transistors. When you have a blown transistor, particularly a driver or output, good to check resistors attached to them. The current draw when the transistor failed may burn them out. Not enough to get them burned looking but when you measure them they may be a significantly higher value. I had a blown driver in Yamaha amp that was a cheap Goodwill find (CA-610II for $20) they used a fusible resistor between it and the output transistor, not noted as such on schematic, but when new transistor did not get channel sounding right found that the emitter resistor had a value thousands of times what it should but looked perfectly normal. Also had redone the thermal paste on the outputs and the one the blown driver had been attached to got some paste on the screws isolating the output transistor collector from power supply leading to a repeat of the blown driver and emitter resistor. Sometimes they are bad, sometimes you will screw up and will be like Charlie Brown when Lucy yanks the football.
I have several pieces nearly the same age to 5 years younger and several have required replacement transistors. A number of transistor designs used in the late 60s into the 70s have a history of noise and other issues and some have such a bad reputation it is recommended to replace them unless they are in a tuner circuit and only replace those if they go bad. The blocky 2SC458LG (made by Hitachi I believe) is one of the most notorious used in a lot of old Sansui and other makes. My project 1970 SX-9000 has a lot of those inside. Some with age just go south and start getting noisy if they don't outright die. One bad one on the front part of an amp channel on my '69-71 Scott would not let the DC offset adjuster on one channel go below 200 mV. New transistor pair fixed it and sounds good with DC near zero.
I will be getting my Onkyo off the work table tonight I can slip my 3000A on there before the tuner I was going to work on, pop the hood and look at the schematic and translate. I will have to look to confirm but if the current at the fuses they talk about go through those resistors you multiple the current by the resistor value to get a mV value to measure across the resistor for the bias adjustment. Conrad Hoffman gives an alternative to Sansui's manual for adjusting the 2000A, especially if you have done his mods to one, and you measure voltage at the four resistors attached to the outputs. They say do the bias adjustment with no load, which is different than Sansui states. When it comes to the DC offset do that with a load attached. They indicate, and based on past experience probably right, it will take at least two rounds to get both close.
When you deal with old vintage items often you will find bad transistors. When you have a blown transistor, particularly a driver or output, good to check resistors attached to them. The current draw when the transistor failed may burn them out. Not enough to get them burned looking but when you measure them they may be a significantly higher value. I had a blown driver in Yamaha amp that was a cheap Goodwill find (CA-610II for $20) they used a fusible resistor between it and the output transistor, not noted as such on schematic, but when new transistor did not get channel sounding right found that the emitter resistor had a value thousands of times what it should but looked perfectly normal. Also had redone the thermal paste on the outputs and the one the blown driver had been attached to got some paste on the screws isolating the output transistor collector from power supply leading to a repeat of the blown driver and emitter resistor. Sometimes they are bad, sometimes you will screw up and will be like Charlie Brown when Lucy yanks the football.
I have several pieces nearly the same age to 5 years younger and several have required replacement transistors. A number of transistor designs used in the late 60s into the 70s have a history of noise and other issues and some have such a bad reputation it is recommended to replace them unless they are in a tuner circuit and only replace those if they go bad. The blocky 2SC458LG (made by Hitachi I believe) is one of the most notorious used in a lot of old Sansui and other makes. My project 1970 SX-9000 has a lot of those inside. Some with age just go south and start getting noisy if they don't outright die. One bad one on the front part of an amp channel on my '69-71 Scott would not let the DC offset adjuster on one channel go below 200 mV. New transistor pair fixed it and sounds good with DC near zero.
