Hitachi HMA-7500 MKII distortion

[forgonati]

Hi guys! I need some help/guidance with this amplifier. I got it cheap together with it's premp HCA-7500, which is working fine. But the power amp has weird distortion when i turn the volume above the certain level, sometimes it's so violent, that even the protection is kicking in and the relays disengaging. I made a video about it, you can check it out.

The previous owner already did a full recap on it, basically every capacitor has been replaced, including the main filter caps. He replaced the bias pots. The driver transistors Q757 and Q758 are also new. What i managed to discover yet:

The distortion appears on both channels. That is weird since the amp is a dual mono construction, basically everything is doubled, transformers, rectifiers, protection circuits, ect.
It's definitely coming from the HMA-7500, i have already tried another source, so the HCA-7500 preamp can be ruled out.
No matter if i use direct coupled or cap coupled input.
No matter if the switch is set to 4 Ohm or 8 Ohm.
If i push BTL button, the LED lights up, i hear relay click and i get no sound. I'm not sure how this BTL (which i believe is a bridged mode) supposed to work...
The main voltages are rock stable, i do have +56 and -56 volts, even when the distortion happens, the voltages are steady.
Voltages on the IC711 (which is some kind of preamp i guess) are there and stable.
I checked every solder joint for cracked solders, the suspicious joints were reflowed.
No loose cable connections.
I did visually check every component, every resistor for cracks or burn signs, i could find nothing.

I will attach some photos and the schematic here, unfortunately it is way too complex to me, i don't really understand how it works, and what are LN, LNF, LP, LD for instance. I did measure the signal on these points and the distortion is there which tells me that the problem might be somewhere before the power stage section.
It's still not clear to me why do i have it on both channels. What is common in this amp that could effect both channels??

Any help would be nice!

 

[chazix]

There's no particular reason to think that this could explain the symptom, but the fusible resistors (shown as black bars in the schematic) should be checked for proper value, or just replaced on suspicion. They are prone to failure.

 

[Goldie99]

I've had massive (unplayable) distortion only from failing fusibles in the past, and wondered the same. It's certainly well worth checking esp. as most can probably be checked in situ.

 

[chazix]

Looking a little more at the schematic, another thing to check is the -12.5V power supply rail (at the emitter of Q802), as this is used by both L and R input stages. (Would also be wise to check the +12.5 rail at Q801's emitter.)

I should add that, even if the above leads somewhere, you would still want to check on the fusible resistors - they are quite likely to fail at some point.

 

[forgonati]

Thanks guys, i will check these voltages. I see that resistors R722, 725, 726, 759, 760, 763, 764, 801 are replaced already. Maybe some other resistors too, but these are definitely new. Probably the previous owner been trying to repair this at some point and finally gave up... Anyway i will check all the other resistors, if i can find any suspicious complonent i will replace them. And i will definitely check those voltages.

 

[Mike_S]

Have a 7500 MKI and it sounds wonderful! Fuse resistors as stated, there are so many in Hitachi's! Yes, they cause most of the issues.
Also make sure the offset and bias are set correctly, on the MKI it's quite the process.

 

[forgonati]

Have a 7500 MKI and it sounds wonderful! Fuse resistors as stated, there are so many in Hitachi's! Yes, they cause most of the issues.
Also make sure the offset and bias are set correctly, on the MKI it's quite the process.

The bias ist set to 35mV, really it's a bit complicated with those 2 trimpots per channel. And the adjustment isn't really well explained in the service manual, it's super confusing. Probably the translation went wrong or something... It mentions trimpot 1 and 2 but you have no idea which is trimpot 1 and 2, on the picture is not shown. Anyway i can measure 35mV on those testpoints R765 and R779 (in series with 1kOhm resistor as per the manual) so i think the bias is more or less OK. By the way the amp runs pretty warm even when idling, but i have read on other forums that this is normal by these amps...

Do you guys understand this circuit? Is here someone who can explain how it works? What are those RN, RNF, NF, RP, RD, this is a bit too much and way above my knowledge, how this amp works... Also i don't know how the bridge mopde should work, as i wrote if i press BTL button, the red LED comes on, i hear relay click but no sound. Should i connect the speaker differently for bridged mode, or when i press it i should get sound out of it just louder? If so, than something is wrong with that bridged mode switch...

 

[chazix]

Do you guys understand this circuit?

I'm not prepared to go that far, but I think I can say that the overall power amp circuit is split across boards, one containing the differential and voltage gain stages, and the other containing a "follower" that has enough current gain to drive speakers. (Both boards also contain some other functions, but I'll stick to the normal-operation audio signal stuff.)

I think I can also speak to the connections labeled with (L or R, left or right channel) followed by N, P, NF, or D:

• The N and P are outputs of the differential/voltage gain board. In normal operation, they carry the same audio (AC) voltage, but have average DC voltages that differ by an amount that helps to produce the right idle current in the output devices on the follower board.
• The NF and D connections both come from the output point of the follower board. They are used to complete two "global" negative feedback paths around the overall power amplifier circuitry. The NF path deals with audio (AC) signals, and the D is for the "DC servo" path.

(I can't speak to the connection that is just labeled "NF" without an L or an R, at least not at a glance. Or even two glances.)

The above is all pretty heavily summarized, but one thing that might help you with troubleshooting: In normal operation, a channel's N, P, NF, and D connections should all have the same audio (AC) signal. So (as I think you observed above) if the N and P signals are distorted in the same way as the NF and D (and speaker output) signals, then the follower board is doing what it should, and the fault is probably in the differential/voltage gain board.

Whew, too much overconfident armchair-quarterbacking! I'd best take a break. I'll let someone else tackle the bridge mode inquiry, but if you can find an owner's manual, it should make that clear.

 

[forgonati]

I'm not prepared to go that far, but I think I can say that the overall power amp circuit is split across boards, one containing the differential and voltage gain stages, and the other containing a "follower" that has enough current gain to drive speakers. (Both boards also contain some other functions, but I'll stick to the normal-operation audio signal stuff.)

I think I can also speak to the connections labeled with (L or R, left or right channel) followed by N, P, NF, or D:

• The N and P are outputs of the differential/voltage gain board. In normal operation, they carry the same audio (AC) voltage, but have average DC voltages that differ by an amount that helps to produce the right idle current in the output devices on the follower board.
• The NF and D connections both come from the output point of the follower board. They are used to complete two "global" negative feedback paths around the overall power amplifier circuitry. The NF path deals with audio (AC) signals, and the D is for the "DC servo" path.

(I can't speak to the connection that is just labeled "NF" without an L or an R, at least not at a glance. Or even two glances.)

The above is all pretty heavily summarized, but one thing that might help you with troubleshooting: In normal operation, a channel's N, P, NF, and D connections should all have the same audio (AC) signal. So (as I think you observed above) if the N and P signals are distorted in the same way as the NF and D (and speaker output) signals, then the follower board is doing what it should, and the fault is probably in the differential/voltage gain board.

Whew, too much overconfident armchair-quarterbacking! I'd best take a break. I'll let someone else tackle the bridge mode inquiry, but if you can find an owner's manual, it should make that clear.

Damn, thank you for this info, and the effort, much appreciated! A lot to learn. Thanks man.

 

[forgonati]

So i had time to poke around in the amp, and every fusible resistor has been already replaced. I did check the voltages on Q801 and Q802, all the voltages are there. +/- 13V at their base, +/- 26V at their collector, +/- 13V at their emitter.
What else should i check? Any hints?

 

[forgonati]

I'm not prepared to go that far, but I think I can say that the overall power amp circuit is split across boards, one containing the differential and voltage gain stages, and the other containing a "follower" that has enough current gain to drive speakers. (Both boards also contain some other functions, but I'll stick to the normal-operation audio signal stuff.)

I think I can also speak to the connections labeled with (L or R, left or right channel) followed by N, P, NF, or D:

• The N and P are outputs of the differential/voltage gain board. In normal operation, they carry the same audio (AC) voltage, but have average DC voltages that differ by an amount that helps to produce the right idle current in the output devices on the follower board.
• The NF and D connections both come from the output point of the follower board. They are used to complete two "global" negative feedback paths around the overall power amplifier circuitry. The NF path deals with audio (AC) signals, and the D is for the "DC servo" path.

(I can't speak to the connection that is just labeled "NF" without an L or an R, at least not at a glance. Or even two glances.)

The above is all pretty heavily summarized, but one thing that might help you with troubleshooting: In normal operation, a channel's N, P, NF, and D connections should all have the same audio (AC) signal. So (as I think you observed above) if the N and P signals are distorted in the same way as the NF and D (and speaker output) signals, then the follower board is doing what it should, and the fault is probably in the differential/voltage gain board.

Whew, too much overconfident armchair-quarterbacking! I'd best take a break. I'll let someone else tackle the bridge mode inquiry, but if you can find an owner's manual, it should make that clear.

N and P has the same distortion (as far i can tell) and the return NF and D carrying it too but with much higher gain. I still don't understand why it appears on both channels.
What i discovered, if i disconnect the speaker, so no load on the output, i cannot see any distortion on the scope anymore. Does this tell you something?

 

[chazix]

N and P has the same distortion (as far i can tell) and the return NF and D carrying it too but with much higher gain.

If this is so, then I am wrong about the output board being a voltage follower, and that makes all my reasoning suspect. Just to confirm: you see a larger audio signal at NF and D than you see at N and P?

no load on the output, i cannot see any distortion

I think that could be explained by the MOSFETs having gone open-channel (but now I'm doubting my understanding of the circuit). When you did the bias adjustment, did that seem to work as expected?

 

[forgonati]

If this is so, then I am wrong about the output board being a voltage follower, and that makes all my reasoning suspect. Just to confirm: you see a larger audio signal at NF and D than you see at N and P?



I think that could be explained by the MOSFETs having gone open-channel (but now I'm doubting my understanding of the circuit). When you did the bias adjustment, did that seem to work as expected?

Sorry if my post was misleading, maybe because english is my third language...
So, the signal level on P and N is the same as on D and NF. But it seems that the distortion is much more violent on D and NF.

The power mosfets were tested, they are OK. I even cleaned their connection to the board (they using terminals to connect them to the amp board, stupid design in my opinion, what if those connectors are getting corroded?)... I didn't touch the bias pots, they were replaced by the previous owner and since i could measure around 35mV at the test points, i saw no reason to mess around with them. The procedure is very strange in the service manual and it's not straight forward... And if i get the distortion on the P and N, doesn't that mean that it comes from the voltage gain board?

It isn't clear why the distortion disappears when no load...

 

[chazix]

I wonder if I'm misunderstanding the nature of the distortion. I confess that I did not view your video before now. (I avoid embedded video as a general rule.) Now that I view the video, I don't know what to make of it. Do you have something you can use as a sine wave signal generator to use as input to the amp, such that you could show what happens to a sine wave?

 

[KeithTr]

Have a 7500 MKI and it sounds wonderful! Fuse resistors as stated, there are so many in Hitachi's! Yes, they cause most of the issues.
Also make sure the offset and bias are set correctly, on the MKI it's quite the process.

My solution (Mark I, not Mark II) after frying some components the very first time, not realizing the filter caps did not discharge) was to replace the jumpers used for checking bias to the back side of the board, where it is much easier to cut the jumper and attach leads and then either replace the jumper or solder the cut ends together. I also added bleeder resistors to the filter caps. One of my favorite amplifiers.

 

[Mike_S]

The bias ist set to 35mV, really it's a bit complicated with those 2 trimpots per channel. And the adjustment isn't really well explained in the service manual, it's super confusing. Probably the translation went wrong or something... It mentions trimpot 1 and 2 but you have no idea which is trimpot 1 and 2, on the picture is not shown. Anyway i can measure 35mV on those testpoints R765 and R779 (in series with 1kOhm resistor as per the manual) so i think the bias is more or less OK. By the way the amp runs pretty warm even when idling, but i have read on other forums that this is normal by these amps...

Do you guys understand this circuit? Is here someone who can explain how it works? What are those RN, RNF, NF, RP, RD, this is a bit too much and way above my knowledge, how this amp works... Also i don't know how the bridge mopde should work, as i wrote if i press BTL button, the red LED comes on, i hear relay click but no sound. Should i connect the speaker differently for bridged mode, or when i press it i should get sound out of it just louder? If so, than something is wrong with that bridged mode switch...

yeah, it's very difficult. Sorry, don't have any answers. Both my 7500 & 8500 mkII do not run warm at all. The only time they did, was when the bias was off.

After it's warmed up, what are the voltages from the speaker terminals with no input on the amp? You might need to adjust things off the speaker terminals to get proper bias set.

 

[forgonati]

I wonder if I'm misunderstanding the nature of the distortion. I confess that I did not view your video before now. (I avoid embedded video as a general rule.) Now that I view the video, I don't know what to make of it. Do you have something you can use as a sine wave signal generator to use as input to the amp, such that you could show what happens to a sine wave?

My Owon does have a signal generator, and i also have a little chinese scope, it's like a toy but it will do the job i guess. Here is an another video about the issue. 1 kHz signal, and with the speaker disconnected (simply with speaker A button) i can go up to 1.70V before the distortion appears. With speaker connected i cannot go up to 0.1V, at 0.08V comes the distortion, at 0.09V it goes copletely crazy and above 0.1V the protection shuts the relays down. I have tried to put this new video into my first post instead of my previous video but i cannot edit that post anymore.

 

[forgonati]

yeah, it's very difficult. Sorry, don't have any answers. Both my 7500 & 8500 mkII do not run warm at all. The only time they did, was when the bias was off.

After it's warmed up, what are the voltages from the speaker terminals with no input on the amp? You might need to adjust things off the speaker terminals to get proper bias set.

You mean the DC offset? Not worrying at all, 25mV on one channel and i think it was 30mV on the other channel. My amp gets warm when idling but if i check the voltages across those 2 resistors like the manual says (through 1kOhm resistor) i get 35mV i think that is fine. Honestly i don't really get the adjusting procedure, it's written very strangely. If i try to turn R757 the reading barely changes, when i turn R758 the voltage changes a lot. But i did not mess with it too much since the reading on the tespoints seems to be fine, it's only 2mV above that 33mV which the service manual says. That shouldn't be a problem at all...

 

[chazix]

1 kHz signal,

Wow. To say the least, that is not the sort of distortion I had imagined. I had assumed that there was just a non-linearity in the amp's response, but this seems more like the amp is breaking into oscillation at points that are at least loosely correlated with the phase of the input signal.

Truthfully, I have no solid ideas. You should double-check that all connections, especially ground connections, are good - but I gather that you've already checked that. You might also try tapping the relays at the onset of symptoms, to see if the symptoms change.

I think one thing I would try is to disconnect the collectors of Q759 and 762, which are meant to detect excessive current in the output stage and, if detected, disable the signal drive to the output stages. It would make sense to first check the output MOSFETs' source resistors, the 0.22 ohm parts. (You won't be able to accurately measure such low resistance values, but they should measure as less than 1 ohm.) The big flaw with this line of thinking is that it seems unlikely for an over-current detection problem to affect both channels.

 

[chazix]

You mentioned the driver transistors having been replaced. Was that done to both channels? How confident are you that the replacements were appropriate?