UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Audio (record players, hi-fi etc)

Notices

Vintage Audio (record players, hi-fi etc) Amplifiers, speakers, gramophones and other audio equipment.

Closed Thread
 
Thread Tools
Old 11th Nov 2021, 11:25 am   #1
MelJon66
Pentode
 
Join Date: Apr 2020
Location: Somerset, UK.
Posts: 173
Default Hitachi HA-250 / HA-330 fusible resistors

Greetings All,

I bought a known faulty Hitachi HA-250 amplifier very cheaply as a source of spares for my more powerful HA-5300 of the same era. The knobs and switches etc. are the same. It’s a low powered bottom of the range Japanese vintage model from 1977 but has a nice solid aluminium look. The HA-330 and possibly other models higher up the range are virtually the same but higher powered, hence the reference in the title.

On receipt it was definitely not working but all the power supply voltages were present. The speaker relay was not clicking in and the right channel heatsink was getting very hot. It was otherwise in nice condition and too good to strip down to leave the rest on a shelf for spares. A few quick checks revealed that the power transistors tested fine (in circuit) but most of the fusible resistors in the power amp section had gone very high or open which seems to be a common problem in these old Hitachi’s. I was tempted to replace them with some standard resistors but finally decided to buy some new fusibles, all 250mA rated, which are still available but some values are very hard to get now. The new resistors cost more than the amp but it may turn out to be a good decision.

Referring to the attached schematic (not great but readable) I changed all the 120R and 47R fusibles – R711, R714, R722 and R723 in both channels. All the 10R ones measured well so I have left them in for now. When I powered up all seemed well for a while. I was able to set the quiescent current across R720L on the left channel but I got no voltage drop across R720R. The speaker relay did not click in. When checking the right channel I noted that R723R had gone open. It didn’t take long for R723L to go the same way. Further checks showed that the collector voltages on all the final transistors Q706 and Q707 were at -28.9V rather than 0V. My initial suspicions were that the Q707’s had shorted emitter to collector or C710 was leaking DC so I removed both from the right channel. Q707R tests fine in diode test mode and C710R reads 1000pf as it should and high DC resistance. C710's are mylar so I get no ESR reading. The emitters of the Q705’s read around 0V and they both test OK in circuit so the large negative voltage is not coming from there.

The fused resistors R723L and R723R are both in the positive half of the push pull cycle and I suspect they have failed because the -28.9V at the collectors of the Q706’s results in a Vce of -60V (effectively double what it should be but within the maximum for a 2SA626 transistor) which will draw more current through the base of Q706 and in turn through the R723 resistor at the output of Q704. Please correct me if I am wrong. So, hopefully the fusibles have saved the transistors. New transistors are expensive and even more so because I believe these are Sziklai pairs so would have to be matched and changed as pairs.

So, my questions are:

1) Are there more thorough tests I can do to be more certain that the Q707’s and C710 are good?

2) What other failure modes would result in -28.9V at the collectors of Q706/Q707?

Any advice appreciated.

Cheers
Mel
Attached Files
File Type: pdf hitachi_ha-250_schematic.pdf (867.1 KB, 110 views)
MelJon66 is offline  
Old 11th Nov 2021, 8:38 pm   #2
Marconi_MPT4
Heptode
 
Marconi_MPT4's Avatar
 
Join Date: Mar 2011
Location: Dorset, UK.
Posts: 519
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Hi Mel,
A fault condition in the output stage which results in an excessive current through R720 (0.22Ω) will produce a sufficient gate voltage via CR902/3 to fire thyristor SCR901. When activated the reference voltage on CR802 for the 14V regulator circuit is biassed to -32V via R912 (2.7k) then clamping reference to -0.6V. Q701~704 lose their positive supply and -29V or so appears on Q706,707 collectors. The speaker protection circuit will activate. Note that thyristor SCR901 is shared with both channels so a fault in one channel will disable both.

If the transistors check out OK, the problem may be due to the quiescent current bias circuit. Favourite here is double diode CR701. One failure mode is to go high resistance or open circuit which will cause excessive current to flow through Q704, Q705 and associated (fuse) resistors. Faulty preset pot is possible.

Is the 14V supply present?

Rich
__________________
To an optimist a glass is half full; a pessimist half empty; an engineer twice as big as need be!
Marconi_MPT4 is offline  
Old 11th Nov 2021, 9:17 pm   #3
Whaam68
Hexode
 
Join Date: Feb 2007
Location: Luton, Bedfordshire, UK.
Posts: 454
Default Re: Hitachi HA-250 / HA-330 fusible resistors

I can’t help with the technical query but it’s worth persevering with the HA-250. Mine is in use daily in my home office and with the loudness circuit engaged it’s a really nice sounding amp.
Whaam68 is offline  
Old 11th Nov 2021, 9:56 pm   #4
MelJon66
Pentode
 
Join Date: Apr 2020
Location: Somerset, UK.
Posts: 173
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Thanks very much for the input Rich,

I have some components out at the moment so can't power on. I will replace them in the morning and see if the 14V supply is present.

When I was unable to set the quiescent current on the right channel I did look closely at the bias adjustment section and noted that CR701R should be bonded to Q705R but had come away. I wondered if Q705R had got hot but it measures fine in circuit. I measure 0.940V forward and open in reverse for CR701R and CR701L which seems ok. I measure 21 ohms resistance across TH701 in parallel with R709 on both channels which also seems ok. The preset pots on both channels seem to operate smoothly and the resistances read similarly as they are adjusted. There does not appear to anything open in either bias circuit but I will check more thoroughly tomorrow.

I did read somewhere that SCR901 can fail and trigger the protection circuit so I will have a look at that too. One post suggested removing it but I am reluctant to do that.

Thanks again.

Cheers
Mel
MelJon66 is offline  
Old 11th Nov 2021, 10:01 pm   #5
MelJon66
Pentode
 
Join Date: Apr 2020
Location: Somerset, UK.
Posts: 173
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Hi Whaam68,

I share your opinion. The HA-250 is incredibly well built for a 20 watt amp and looks very chunky. There is no sign of anyone having meddled with it and no sign of heat stress, despite its years. I am looking forward to getting it up and running.

Cheers
Mel
MelJon66 is offline  
Old 12th Nov 2021, 5:04 pm   #6
MelJon66
Pentode
 
Join Date: Apr 2020
Location: Somerset, UK.
Posts: 173
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Hi Rich and All,

I have now re-assembled the HA-250 but have NOT replaced the fused R723L and R723R resistors. I am reluctant to replace these and risk them and other fusibles blowing again and potentially damaging the transistors.

When powering up under these conditions the protection circuit is not triggered. Not surprising perhaps because the Q704's are not driving the Q706's.

In the power supply section Q801's emitter is at 13.28V and Q802's is at -13.3V. These seem a little low but close enough. Maybe our mains voltage is lower than it was in the 1970's. Q801 and Q802 test fine in circuit.

The voltages on SCR901 are as follows:
Gate: -32.19V
Anode: +14.49V
Cathode: -32.20V

The voltage at R912 - furthest from SCR901 anode = +14.6V

Re-checking the resistances of the components in the bias circuit:

VR701L and VR701R = 93 ohms on both channels
TH701 in parallel with R709 = 21 ohms on both channels
Forward voltage drop across CR701L = 0.934 and for CR701R = 0.936

So, all seems well in the bias circuit.

I do suspect that the current overload on the R723's must be marginal. I was able to easily set the quiescent current on the left channel before R723L fused. By the time I had got to the RH channel the R723's had fused. I fitted 250mA rated fusibles as the schematic doesn't specify higher ratings. It seems too rash to stick higher rated fusibles in until I have eliminated any possibility of a fault.

With the R723's removed is there anything else I can usefully test?

Rich, when you said "Q701~Q704 (Q703?) lose their positive supply and -29V or so appears on Q706,707 collectors" I can follow how the positive supply is lost but not how -29V appears on the collectors. Please can you explain?

Cheers
Mel
MelJon66 is offline  
Old 13th Nov 2021, 9:36 am   #7
MelJon66
Pentode
 
Join Date: Apr 2020
Location: Somerset, UK.
Posts: 173
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Hi All,

I think I have identified the cause of the initial fault. This job was about replacing the fusible resistors that had gone high or open. Basic checks didn't show any other faults so I just replaced the faulty resistors. Having done that and before I switched on I checked for shorts and did a visual check and noticed that the bias pots VR701L and VR701R were in very different rotational positions. I thought this was suspicious so I set them both to minimum (fully anticlockwise) to achieve minimum bias, thinking that was the safest starting position. That was a mistake. I woke up at 3am this morning realising this. The unconscious workings of the brain are a mystery but it would be better if these thoughts came to me during waking hours and preferably BEFORE I make the mistakes .

These pots operate very smoothly from around 35 ohms up to about 250 ohms. BUT, my subsequent checks show that there is a very rough spot just at the fully anticlockwise position where the resistance jumps to over 200 ohms. So, I have inadvertently set the pot resistance high and applied too much bias voltage and current. After the event I realise that setting the minimum resistance of a pot using a meter is definitely best practice! Rookie error .

This fault is consistent with what I noted when I switched on. I was able to set the Q current for the left channel but by the time I got to the right channel resistors R723 had fused. The temperatures of Q704 and Q705 were probably rising at that time but I was only monitoring the temperatures of the output transistors Q706 and Q707. It was all happening in slow motion but I still missed it.

The Q704’s and Q705’s still measure ok in circuit so perhaps the fusible resistors have saved them. I will replace the fusibles but before I switch on I would like to be sure there isn’t another fault that is causing the -28.9V on the collectors of Q706 and Q707. I can’t imagine that replacing the fusibles will fix that, but I might be wrong! The protection circuit is not activated when I switch on now (with R723's out of circuit) and I measure +13V ish at the emitter of Q801. It may have been activated by the original fault but has been reset. For now the reason for the -28.9V remains a mystery.

Any suggestions?

Cheers
Mel

Last edited by MelJon66; 13th Nov 2021 at 9:47 am.
MelJon66 is offline  
Old 13th Nov 2021, 10:27 am   #8
Whaam68
Hexode
 
Join Date: Feb 2007
Location: Luton, Bedfordshire, UK.
Posts: 454
Default Re: Hitachi HA-250 / HA-330 fusible resistors

I’m paranoid and not having anything more sophisticated always power up after working on something using a lamp limiter which has saved my bacon a few times and gives an instant clue the amp is drawing too much current if the bulb glows brightly. Might be worth knocking one up? Mine is very crude but does the job.
Mike
Whaam68 is offline  
Old 13th Nov 2021, 11:19 am   #9
MelJon66
Pentode
 
Join Date: Apr 2020
Location: Somerset, UK.
Posts: 173
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Thanks Mike,

I initially brought it up slowly with a dim bulb tester feeding into a variac. The bulb did brighten slightly as the voltage increased but there was nothing to be concerned about so I disconnected the dim bulb and variac. Inadvertently finding a high resistance position on the bias pot almost certainly caused too high a bias current. I now recall someone saying that the dim bulb should be left in circuit while you adjust the bias for that very reason. Also, the thermistor that could have compensated for the high bias was not bonded to the transistor as it should have been so it probably reacted slowly, if at all.

Hindsight is a great thing but not as wonderful as foresight .

I'd still like to get the HA-250 going but if the transistors have blown they are too expensive to replace. I am still open to suggestions on how I might proceed.

The good thing is that I have to repeat this exercise with my HA-5300 and I will not make the same mistakes again. I'm sure I'll discover some new ones but that's life!

Cheers
Mel
MelJon66 is offline  
Old 14th Nov 2021, 11:15 am   #10
MelJon66
Pentode
 
Join Date: Apr 2020
Location: Somerset, UK.
Posts: 173
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Hi All,

Being fairly sure that I understood why the fusible resistors had blown the last time I switched on and having replaced them again and checked for shorts I felt I had done as much as I could. Accepting that I might never fully understand why there was -28.9V on the 0V rail when the fusible resistors were out I decided to bite the bullet and try again. With dim bulb in circuit (note the learning point there!) I switched on, the dim bulb remained resolutely dim and the speaker relay clicked. Eureka!

The 0V line is now at 0V so the -28.9V appears to have been a bit of a red herring. Hitachi describe this amp as having an “Inverted Darlington Circuit”. This appears to be the same as a Sziklai pair or Complementary Feedback Pair. At risk of talking tosh I would guess that when the R723 resistors blew it effectively broke the links between the Siklai pairs (Q704 and Q706) on the positive side of the push pull. Q704 apparently provides the bias for Q706 which was now missing so the positive half of the push pull wasn’t doing anything. Hence the negative half dominates and puts a negative voltage on the 0V line. Anyway, that’s my best uneducated guess.

The HA-250 manual does not mention bias adjustment at all but the one for the HA-330, which has the same circuitry, specifies 8mV +/- 4mV across R720 to give a nominal 40mA at idle. 8mV was easily achieved as the bias pots are smooth and not particularly sensitive. It now idles as cool as a cucumber compared with the fiery little radiator it was in its non-working condition.

As I noted in a previous post one of the thermistors had come away from the transistor (Q705) that it should have been bonded to. I put some heatsink compound between them and wrapped them with a small tie wrap to hold them together. It seems to work but is there a recommended method for bonding them? The bonding is probably important as the Q705’s run hotter (30 degrees at idle) than the Q706’s and Q707’s that are fitted to the large heatsink (24 degrees at idle). I imagine this might change when the amp is working hard but all the thermal bias compensation appears to be made via the thermistors bonded to the Q705’s.

My next step is to put some waveforms through it and see how it performs.

So far so good.

Cheers
Mel
MelJon66 is offline  
Old 15th Nov 2021, 3:10 pm   #11
Whaam68
Hexode
 
Join Date: Feb 2007
Location: Luton, Bedfordshire, UK.
Posts: 454
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Glad to hear you got it working!
Interesting yours is running cool as while sounding really good mine runs fairly warm... noticeably so once it's been running a while. Might be time for a check over...
Mike
Whaam68 is offline  
Old 16th Nov 2021, 7:01 pm   #12
MelJon66
Pentode
 
Join Date: Apr 2020
Location: Somerset, UK.
Posts: 173
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Hi Mike,

Yes, the right channel was running hotter than the left at around 40C before the fusibles were changed. My HA-5300 is just the same with one channel much hotter than the other. The HA-5300 still works in this condition but I haven't used for a long time, not wanting to risk it going pop. The HA-250 had already stopped working when I got it. Yours might go the same way if you leave it too long. It's an easy job if you don't make the rookie mistakes I did .

There are 14 fusibles with 3 different values. I have changed 8 of them so far, leaving the 10R values until later because they all measure well. When you get to the point of doing the work let me know and I will give you the details. If this thread is closed by then just PM me.

Cheers
Mel
MelJon66 is offline  
Old 16th Nov 2021, 7:56 pm   #13
Whaam68
Hexode
 
Join Date: Feb 2007
Location: Luton, Bedfordshire, UK.
Posts: 454
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Thanks Mel thats kind will do. Prob at the mo is getting the time!
Whaam68 is offline  
Old 20th Jan 2022, 2:28 pm   #14
MelJon66
Pentode
 
Join Date: Apr 2020
Location: Somerset, UK.
Posts: 173
Smile Re: Hitachi HA-250 / HA-330 fusible resistors

Greetings All,

Just a quick wrap up on this thread.

I did eventually complete the fusible resistor replacements by changing all the 10R values, despite them measuring well. With the work done the HA-250 amp was working perfectly. Sine wave input tests with an oscilloscope across 8 ohm load resistors confirmed that output level was linear with no visible gliches in the output trace. It's a 20 watt amp and it produces 30 watts with a 1KHz sine wave input so I was very happy with that. To try and head off any future problems with failing electrolytic caps I did some checks on the capacitance values and ESR and they were all fine so I have left well alone. These Hitachi amps are very well built with top quality Japanese Nichicon caps and they obviously last well. I am very impressed with the build quality (not a dry joint anywhere and no sign of heat stress), ease of access and the quality of the switches and components used, apart from the fusible resistors which appear to be their achilles heel. You just don't get this quality on low end products these days.

I have been using it in my study connected to my PC using MusicBee or Boom Radio as input and it sounds excellent to my ears. I don't remember reading any reviews of this model back in the 70's/80's as it was a bottom of the range model and the top of the range Hitachi mosfet amps seemed to get all the attention. As Mike (Whaam68) said, the loudness button does a great job at low listening levels and I use it all the time. The loudness button on this produces a better balanced sound than the larger Hitachi HA-5300 (Toooo much bass!) which I have also refurbished and will report on soon. Loudness buttons are frowned on by HiFi purists but who cares when it sounds good. One problem I found was that the amp picks up hard drive and mouse movement sounds from the PC so I fitted a cheap ground loop isolator (Bumper brand) on the input sockets and it's now silent with no loss of sound quality to my ears.

So, a great outcome for a cheap "spares or repair" purchase. I'm enjoying using it more than other amps I've got that cost ten times the price . Not bad for a lowly 20 watt amp with over 40 years on the clock.

Cheers
Mel
MelJon66 is offline  
Old 21st Jan 2022, 7:46 am   #15
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,639
Default Re: Hitachi HA-250 / HA-330 fusible resistors

Well done,glad you got it fixed. the bias pots on all these vintage amps are worth replacing with 20 turn modern presets, makes setting the bias loads easier.

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 3:23 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.