Maybe your patience is less than mine.
I can get steady Iq with a constant voltage bias circuit.
Maybe you have smaller heatsinks than mine.
This is puzzling to me.
I have a question: if you measure Vgs across the mosfets, do the Vgs values drop as the devices go from cold to hot?
Tombo do you have Sissy Sit or M25 (I was assuming M25)?
Last edited:
No problems for the measurements if I could borrow De Lorean DMC-12 (there is some thunderstorm here ATM)
I said that Sissy is on my to-do list so no measurements yet. Just thinking in advance how to address some problems.
Aaaaahhhh
You're just contemplating what you might do, if you actually have a problem.
After 20 hour cook and bias I put both sissy’s in my active system. System uses DEQX preamp for active crossover, dac and preamp function. JBL Harsfield horns, Fostex super
Tweeter (driven by one amp and using passive xover with compression driver) and Danley DTS-10 subwoofer.
500-20kHz is handled by one SissySIT.
80-500 by the other SissySIT
<80 by big Behringer amp.
First impression - mids and highs are simply holographic. There is magical fairy dust in the SIT.
After a while I put an F5 in for 80-500hz duty and did level adjustments. The F5 seems to have iron grip on the JBL 150-4C 15” drivers.
I’m going to run this combination for a while and see what I think. I predict a very late listening session tonight.
Tweeter (driven by one amp and using passive xover with compression driver) and Danley DTS-10 subwoofer.
500-20kHz is handled by one SissySIT.
80-500 by the other SissySIT
<80 by big Behringer amp.
First impression - mids and highs are simply holographic. There is magical fairy dust in the SIT.
After a while I put an F5 in for 80-500hz duty and did level adjustments. The F5 seems to have iron grip on the JBL 150-4C 15” drivers.
I’m going to run this combination for a while and see what I think. I predict a very late listening session tonight.
I have the Sissy SIT and all parts were from ZM sans the SIT, so I can rule out the cause due to fake parts.It would seem to me that you might have a situation where you have constant voltages at the gates of the mosfets and one or both of the hall sensors aren't operating correctly.
If both channels have the same mosfets, then both channels should behave the same.
Please don't tell me you bought your mosfets off ebay.
It wouldn't hurt to check every resistor value and every soldered joint on the board.
I checked the PC817 and 7805 and they are operating well.
When I dial down the Iq, it will creep immediately and so will the DC offset. What I will do is to cool down the set and start again with the lid off, it will settle around 154mV with almost perfect offset.
Yeah Sissy Sit is kind of a different animal altogether, I can see how biasing could be possibly hit and miss.
I keep getting confused cause we’re discussing 2 different output stages, and I keep thinking people are referring to M25, which definitely should not cause such problems.
ZM - here you go. The amp rack is top to bottom: F5, sissy x 2, aleph J mono x 2, Behringer. Passive pre on top temporarily for F5. J Mini stereo amp is by my workbench, and F5 #2 is on the floor.
Need to do final tuning in DEQX for levels.
I’m listening to a favorite John Zorn Masada album now. Sounds incredible. Listened to a Haydn quartet, amazing. Albert king, Cars, and more were also wonderful. This has a magical midrange and highs.
Need to do final tuning in DEQX for levels.
I’m listening to a favorite John Zorn Masada album now. Sounds incredible. Listened to a Haydn quartet, amazing. Albert king, Cars, and more were also wonderful. This has a magical midrange and highs.
Attachments
Iron Maiden and Masada. Nice!
Which Masada Album? Abidan from Masada 3 Gimel has been one of my listening-test pieces for years. Love the buildup of the instruments and the contrabass solo towards the end. Have you checked the Book 2 and Book 3 recordings? Some true gems among those.
Mcgillroy, I was listening to Alef, the odd numbered tracks and the 50th Birthday String Trio Vol 1. I have 1, 4,5,7, and 8 from “book 1”, about 1/2 of book 2, and I got the set of book 3. Plus some odd and ends like circle maker, live sets, guitars. I’ve been fortunate to see various incarnations of Masada starting sometime in the 90s at knitting factory, then tonic, and also in SF. One of my top live music experiences for sure. The diversity of the crowd is always impressive: young and old, jazzbos and punks. Other live highlights are Uz Jsme Doma, Dead Can Dance, Desmond Decker, Subhumans, hammers of misfortune, and the almighty (and recently retired) Slayer. Maiden was pretty epic, too.
I think I have managed to contain the creep curse after many weeks. So the investigation started with the thermal coupling with the heatsink as the left channel was ok but not the right. Suspected that it was not screwed down well or the SIT wasn't flushed to the heatsink. The infrared thermometer showed consistent temperature readings on the left and right SIT, so scratch that out.
Attention turned to the 7805 and the PC817 if they are operating within their specs. No wild readings or saturation too. The MOSFET were behaving. But the creep occurs no matter how I dial down the Iq pot. BUT if I were to keep the lid off, the SIT is nice and good.
I decided to put a small fan directed at the heatsink where the SIT is. That kept the Iq fairly stable! That proves that it was the SIT and not other components which were going crazy. The trick is to dial down the Iq pot very significantly such that the value is ~110mV when it turned on for the first 5 minutes. The Iq will move by another 50-60mV as it reaches the thermal equilibrium. The room ambient temperature also plays a part, our weather in Singapore can fluctuate between 27 to 34C, and the Iq does creep further when its in the afternoon.
Will continue to monitor this further before removing the multi-meters. Hopefully, this bit of information will help anyone else who has the same issue.
Attention turned to the 7805 and the PC817 if they are operating within their specs. No wild readings or saturation too. The MOSFET were behaving. But the creep occurs no matter how I dial down the Iq pot. BUT if I were to keep the lid off, the SIT is nice and good.
I decided to put a small fan directed at the heatsink where the SIT is. That kept the Iq fairly stable! That proves that it was the SIT and not other components which were going crazy. The trick is to dial down the Iq pot very significantly such that the value is ~110mV when it turned on for the first 5 minutes. The Iq will move by another 50-60mV as it reaches the thermal equilibrium. The room ambient temperature also plays a part, our weather in Singapore can fluctuate between 27 to 34C, and the Iq does creep further when its in the afternoon.
Will continue to monitor this further before removing the multi-meters. Hopefully, this bit of information will help anyone else who has the same issue.
Last edited:
I believe I found cheap and easy solution for much better Iq stability , both in Temp. and mains voltage (read Urails) domain
have some family things to take care of in next few days and after that will check everything again ; if things stay so rosy** during second check-uo , will post exact tutorial , with few illustrative piccies
cost of parts ....... 3$ approx.
I believe , with some care , everything can be done from top side of pcb , so no need for de-attaching it from htsnk , if you did mount it without L bracket
in case you did use L bracket - even better and easier
everything applies to SissySIT , Babelfish M25 is behaving ...... impeccable
Indra was again da Man ..... always having most intriguing info - he led me in some of threads to take in account optocoupler with Darlington output
though , it was tough beast to tame
**will try of course to make it even rosier
have some family things to take care of in next few days and after that will check everything again ; if things stay so rosy** during second check-uo , will post exact tutorial , with few illustrative piccies
cost of parts ....... 3$ approx.
I believe , with some care , everything can be done from top side of pcb , so no need for de-attaching it from htsnk , if you did mount it without L bracket
in case you did use L bracket - even better and easier
everything applies to SissySIT , Babelfish M25 is behaving ...... impeccable
Indra was again da Man ..... always having most intriguing info - he led me in some of threads to take in account optocoupler with Darlington output
though , it was tough beast to tame
**will try of course to make it even rosier
OK , being in a rush, outa workshop ..... I believe I found solution for that issue
will check in next few days that too
