Hi All,
I have a Jims audio Hiraga amplifier I built a couple of years ago.
Wonderful amp. Runs daily in my main system. Love it.
But it has a problem in ONE channel.
The gain in that channel drops occasionally. If I disconnect the input RCA’s and cycle the power switch it comes good. Works fine. Usually for the rest of the day. Been putting up with the problem for at least a year. Amp is so good otherwise I just live with this issue.
I have an electronics tech background. I’ve replaced many components, but not all, on the board to no avail. Checked and resoldered all joints.
I strongly suspect that this channel is getting parasitic oscillations. It’s unstable at high frequencies. Possibly MHz. It never happens to the other channel. I’ve swapped the boards left and right and the fault follows the suspect board.
Both boards use the SAME plus / minus power supply. Both rails are regulated.
If anyone would know what to do with this board it could be you guys on diyaudio. So I finally thought to ask? Could there be some component slightly out of spec on that board that is not stopping positive oscillations like the other board which doesn’t have the problem? Or is there a harmless remedy?
Any help appreciated.
Regards,
Paul.
I have a Jims audio Hiraga amplifier I built a couple of years ago.
Wonderful amp. Runs daily in my main system. Love it.
But it has a problem in ONE channel.
The gain in that channel drops occasionally. If I disconnect the input RCA’s and cycle the power switch it comes good. Works fine. Usually for the rest of the day. Been putting up with the problem for at least a year. Amp is so good otherwise I just live with this issue.
I have an electronics tech background. I’ve replaced many components, but not all, on the board to no avail. Checked and resoldered all joints.
I strongly suspect that this channel is getting parasitic oscillations. It’s unstable at high frequencies. Possibly MHz. It never happens to the other channel. I’ve swapped the boards left and right and the fault follows the suspect board.
Both boards use the SAME plus / minus power supply. Both rails are regulated.
If anyone would know what to do with this board it could be you guys on diyaudio. So I finally thought to ask? Could there be some component slightly out of spec on that board that is not stopping positive oscillations like the other board which doesn’t have the problem? Or is there a harmless remedy?
Any help appreciated.
Regards,
Paul.
Last edited:
Do you have an oscilloscope to measure if your suspected hf-oscillations are present?
A spectrum analyser would be better.
Why do you disconnect input RCA's and a power cycle?
Does the disconnecting alone stops the sudden gain drop?
Does the power cycle (RCA's remain connected) alone stops it?
It cannot be a loose connection in the feedback resistor (220Ω or 300Ω), but it can the feedback-to-ground resistor (10Ω) with a problem - when the drops occur, is there still some sound? Then it has become a nunity gain amp.
My Hiraga does have a small cap over the feedback resistor to prevent this. It is a 3n3 yielding a rolloff from 230kHz onwards.
A spectrum analyser would be better.
Why do you disconnect input RCA's and a power cycle?
Does the disconnecting alone stops the sudden gain drop?
Does the power cycle (RCA's remain connected) alone stops it?
It cannot be a loose connection in the feedback resistor (220Ω or 300Ω), but it can the feedback-to-ground resistor (10Ω) with a problem - when the drops occur, is there still some sound? Then it has become a nunity gain amp.
My Hiraga does have a small cap over the feedback resistor to prevent this. It is a 3n3 yielding a rolloff from 230kHz onwards.
thanks for the reply!
my scope goes to 20Mhz. Not sure if enough. But want to try your other suggestions first.
Also,
ok what I do to fix the problem is like hocus pocus magic ritual. But it works.
1. Pull the left RCA out. (problem in right channel, just want to hear that one).
2. Pull the right RCA out ONLY enough to disconnect the earth leaving the inner active connected.
[ At this point I can hear crazy high audio freq screaming noise ]
3. Cycle the power. Sometimes takes a few go's. Then gain and volume come back.
4. Replug RCA's.
Disconnecting RCA alone does stop it sometimes but rarely.
Power cycle alone usually does not fix it.
Usually need the whole steps above.
I am going to RE CHECK the feedback divider resistors.
I may up the resistor to 300 not 250 ohm.
I may try the bypass cap 3n3.
I have altered the bias on this amp to UP the steady state current.
It hasn't caused a problem on the other board so I guess it's not behind the problem I have,
BUT, I just found another big Hiraga thread on diyaudio, and it talks of having a bigger resistor for feedback with the higher power amp. Mine is actually the 15W version. I'm running it at about 25W with a 500VA toroid.
thanks,
I'll let you know how it goes.
Paul.
my scope goes to 20Mhz. Not sure if enough. But want to try your other suggestions first.
Also,
ok what I do to fix the problem is like hocus pocus magic ritual. But it works.
1. Pull the left RCA out. (problem in right channel, just want to hear that one).
2. Pull the right RCA out ONLY enough to disconnect the earth leaving the inner active connected.
[ At this point I can hear crazy high audio freq screaming noise ]
3. Cycle the power. Sometimes takes a few go's. Then gain and volume come back.
4. Replug RCA's.
Disconnecting RCA alone does stop it sometimes but rarely.
Power cycle alone usually does not fix it.
Usually need the whole steps above.
I am going to RE CHECK the feedback divider resistors.
I may up the resistor to 300 not 250 ohm.
I may try the bypass cap 3n3.
I have altered the bias on this amp to UP the steady state current.
It hasn't caused a problem on the other board so I guess it's not behind the problem I have,
BUT, I just found another big Hiraga thread on diyaudio, and it talks of having a bigger resistor for feedback with the higher power amp. Mine is actually the 15W version. I'm running it at about 25W with a 500VA toroid.
thanks,
I'll let you know how it goes.
Paul.
Does the amp have any speaker protection? What you are describing sounds like a dirty output relay.
Hiraga's don't need protection (-relay's), there starting silent by the very design.
Mine only has a supply and amp boards in a case with connectors, nothing more. Not even a fuse. Serves several years now.
Mine only has a supply and amp boards in a case with connectors, nothing more. Not even a fuse. Serves several years now.
I'm not familiar with the amp, does it have an output capacitor? What would protect your speaker if it went DC?
Sounds like Loose solder / input cap is going.
Use your scope & 8ohm load across the output, with a 1khz signal on the input, see what it's doing.
Make sure you have a 8ohm -100watt load & heatsink it correctly so you don't burn your resistor out.
Use your scope & 8ohm load across the output, with a 1khz signal on the input, see what it's doing.
Make sure you have a 8ohm -100watt load & heatsink it correctly so you don't burn your resistor out.
It's symmetric (search!), no output cap. It will fry the LS with 1A dc.
Mine is the 8W driving a >50W ESL+sub Final 0.4.
I’ve swapped the boards left and right.
Problem follows bad board. Only one channel faulty.
So the input cap is fine. Thanks.
Pardon me what is the LS?
I have been concerned about no speaker protection.
It’s symmetrical but does that mean my speakers are always safe from say an output device failing?
I just haven’t got around to putting the speaker protection in.
I don't even bother dc protection with a Hiraga, as the design needs no such.
LS = loudspeaker, without incovenience... Most are magnatic, electrostatic variants are more rare (but sounds imho way better).
Nothing granted, your speakers will never be safe regardlessly of whatever the topology proper: it it fails, everything susceptible is cast down too. A 2 'currency symbol' protection circuit will suit your wallet.
So, no worries mate with an Hiraga, 20,000 km's afar!
LS = loudspeaker, without incovenience... Most are magnatic, electrostatic variants are more rare (but sounds imho way better).
Nothing granted, your speakers will never be safe regardlessly of whatever the topology proper: it it fails, everything susceptible is cast down too. A 2 'currency symbol' protection circuit will suit your wallet.
So, no worries mate with an Hiraga, 20,000 km's afar!
Post pic of your board, & BTW the input cap is on the board that you keep swapping.. ( depending on what boards you are using )
Where did the semiconductors come from? Chinese clones?? I'd say it's definitely oscillating. From your fault description it has all the hallmarks.
Instability occurs in the common emitter predrivers due to unsuitable device combination. You should remove any voltage regulators and protection circuits to enhance stability and sound quality.
N101N,
thanks for this. I believe this is getting close.
But regulators improve sound quality. I've tried it on and off.
The hiraga has a low PSRR and unlike a large voltage supplied KRELL for example it benefits from a regulated supply.
Yes it is from Honk Kong Jims Audio. Power trannies and board and most parts seem top quality.
q1, q5 pre drivers ? Should they be replaced? From where? China not trust worthy?
I'll include circuit.
thanks for this. I believe this is getting close.
But regulators improve sound quality. I've tried it on and off.
The hiraga has a low PSRR and unlike a large voltage supplied KRELL for example it benefits from a regulated supply.
Yes it is from Honk Kong Jims Audio. Power trannies and board and most parts seem top quality.
q1, q5 pre drivers ? Should they be replaced? From where? China not trust worthy?
I'll include circuit.
