never use less than 370C , with proper tip ....... and if you're soldering fat wire or something substantial , feel free to crank it all the way up , with intention that your soldering lasts as short you can manage
larger tip means much lesser temp hysteresys on very tip of ..... tip
so , better to use slightly fatter tip than slightly narrower
I have 2 solder stations on my bench , one equipped with 7mm chisel tip type for fat wires and mosfet legs , and second ( in fact that being first/main) with tiny tip .... which isn't at all tiny - not pen/needle sharp , but more like as tiny flat chisel - having flat part of some 0.7mm
that one doing all my SMD presoldering , using hot air only as final soldering operation
so , with that amount of heat , which you most probably are investing in your soldering- even with pads which on eye are staying intact , you're compromising their bond to pcb
so , think where you're doing it wrong
no problem if SIT is close to Donut - look at my builds
larger tip means much lesser temp hysteresys on very tip of ..... tip
so , better to use slightly fatter tip than slightly narrower
I have 2 solder stations on my bench , one equipped with 7mm chisel tip type for fat wires and mosfet legs , and second ( in fact that being first/main) with tiny tip .... which isn't at all tiny - not pen/needle sharp , but more like as tiny flat chisel - having flat part of some 0.7mm
that one doing all my SMD presoldering , using hot air only as final soldering operation
so , with that amount of heat , which you most probably are investing in your soldering- even with pads which on eye are staying intact , you're compromising their bond to pcb
so , think where you're doing it wrong
no problem if SIT is close to Donut - look at my builds
I reheated all solder connections on both boards, front and back. There was at least one bad joint on the board I had a problem with. I cooked up and biased. I’m around 1.8A on the outputs. It’s never rock solid at 1.8, but close enough. Offset is also not able to stay rock solid.
This time I got front end to dial in nicely. 20mA / zero offset.
Put in jumpers, attach sacrificial speakers, play some Emilíana Torrini, then some Eilen Jewell. Beautiful.
DC offset on outputs is higher than I would expect. 200mv+. Is there a step to adjust offset after installing jumpers that I skipped? Or do I just live with it? What is an acceptable value?
The crap part is that I now need to build the 2nd amp to put in my main system. It’s an active system using a DEQX preamp / processor. I need the amp on my 15”s to have same gain as the mid horns / supertweeter amp. And I don’t think SissySIT has same gain as F5.
This time I got front end to dial in nicely. 20mA / zero offset.
Put in jumpers, attach sacrificial speakers, play some Emilíana Torrini, then some Eilen Jewell. Beautiful.
DC offset on outputs is higher than I would expect. 200mv+. Is there a step to adjust offset after installing jumpers that I skipped? Or do I just live with it? What is an acceptable value?
The crap part is that I now need to build the 2nd amp to put in my main system. It’s an active system using a DEQX preamp / processor. I need the amp on my 15”s to have same gain as the mid horns / supertweeter amp. And I don’t think SissySIT has same gain as F5.
if you can't get 0 offset and you're still on positive side, either increase P2+R18 combo or decrease R19
it seems you got tricky combo of Ugs voltageswith your SIT and mosfet
easiest - just solder additional resistor across R19 , piggy-back style
180K will do , to get it to 30K
don't forget to re-bias and re-offset
both Iq and DC offset strict in temp equilibrium , nature of the beast ....... it could be done more precise and stable but no way without resorting to servo(s) and that would be simply too tricky to troubleshoot in case of any of common Greedy Boyz Palette of Mistakes
NB that input buffer DC offset is having nothing with OS DC offset - there is no "after installing jumper" consequence
buffer DC offset is all about not magnetizing autoformer, and there is Silmic+MKC combo afterwards , for DC decouple between FE and OS
so , to recapitulate - you want 0 DC offset on output in temp. equilibrium state and steady mains , simply don't bother thinking about powering up,heating up etc. - it behaves enough even for Lowthers
it seems you got tricky combo of Ugs voltageswith your SIT and mosfet
easiest - just solder additional resistor across R19 , piggy-back style
180K will do , to get it to 30K
don't forget to re-bias and re-offset
both Iq and DC offset strict in temp equilibrium , nature of the beast ....... it could be done more precise and stable but no way without resorting to servo(s) and that would be simply too tricky to troubleshoot in case of any of common Greedy Boyz Palette of Mistakes
NB that input buffer DC offset is having nothing with OS DC offset - there is no "after installing jumper" consequence
buffer DC offset is all about not magnetizing autoformer, and there is Silmic+MKC combo afterwards , for DC decouple between FE and OS
so , to recapitulate - you want 0 DC offset on output in temp. equilibrium state and steady mains , simply don't bother thinking about powering up,heating up etc. - it behaves enough even for Lowthers
I am also having issues with my left channel as I couldn't dial down the climbing offset. Checked on the 7805 as instructed by ZM and also looked at the voltages of pins 3(-3V65) and 4(-4V84) of the optocoupler.
ZM, how much of the offset can I dial down if using 180K resistor to piggyback with R19?
ZM, how much of the offset can I dial down if using 180K resistor to piggyback with R19?
180k piggybacks installed on R19. Cooked and adjusted for 2.5 hours. Still drifting bias and offset. Are there certain critical voltages I should measure and report here to help troubleshoot?
Should I consider replacing mosfets?
Looks like Sits are 3.42v from pras
Should I consider replacing mosfets?
Looks like Sits are 3.42v from pras
Attachments
Well, how much drift do you actually have ?
My SissySIT at its current settings moves between 180W overall power consumption when cold to about 230W when hot ... strong dependence on wall voltage as well, bias could change by close to 10% with a change in wall voltage (corresponding with rail voltage) of 5%.
Accordingly, when I set the DC output offset at zero for, say, a rail voltage that corresponds to 225V wall voltage, it will move between -50mV and +50mV when wall voltage varies 220-230V.
In a discussion with ZenMod, we found that this is within the design parameters of the bias circuit because of the combination of depletion mode and enhancement mode output device.
Best regards, Claas
My SissySIT at its current settings moves between 180W overall power consumption when cold to about 230W when hot ... strong dependence on wall voltage as well, bias could change by close to 10% with a change in wall voltage (corresponding with rail voltage) of 5%.
Accordingly, when I set the DC output offset at zero for, say, a rail voltage that corresponds to 225V wall voltage, it will move between -50mV and +50mV when wall voltage varies 220-230V.
In a discussion with ZenMod, we found that this is within the design parameters of the bias circuit because of the combination of depletion mode and enhancement mode output device.
Best regards, Claas
I got a couple of those non-100ohm buggers in my kit, and of course I soldered one on a board, so I'm going to replace it with the one that has not been soldered yet.
I did find find some 100ohm 1/8 watt metal film resistors. Can I use them on the gate?
Thanks,
Brad
I cooked for 2.5 hours. I was around 48 degC at the heatsink fins. Chassis is Deluxe 4U - standard depth.
All testing below under temp equilibrium.
One channel I'm setting around 1.8A. That stays fairly stable. When I work to zero the offset and let cook for a while (say 15 minutes more) it may drift to 35mV. This is probably OK.
The trouble channel is I set for 1.75A and let cook. I walk back and it is is at 1.95A. I dial it down. Say 1.7. Then it drifts back up after 15 minute cook. I try to zero offset. After 15 minute cook 100mV or 150mv offset drift. repeat the cycle. It never settles to steady Iq or offset.
I have my 2nd set of boards and parts currently not assembled. I can remove a MOSFET from Kit #2 and replace it on trouble channel to see if that makes a difference. I can also do a Vgs check on those 2 loose MOSFETs to pick the best fit.
I thought I might try to switch SIT's between the channels. Or take the spare SIT's for the 2nd amp and switch one in. BUT if I assume Pras matching is good, my SITs are at 3.42V, 3.42V, 3.4V, 3.4V. They are all very close... So this is probably not the best approach?
Another thought was is build the other set of boards (I need to do it anyway), with very close attention to details. Measure each part, find any buggers mixed in, cross off each part on schematic when installing. Then visually compare the 2 boards to ensure all resistors are correct, diodes are correct, etc.
I can also go and measure voltages at certain points to see if that shows something wrong.
Another option, I do have a some matched quads of IRFP9240's on hand. Vgs for these are in range of 3.85 - not sure if that is OK, too high, to low? If I do use 9240's would other values on the SIT board need to change?
Thoughts?
One channel I'm setting around 1.8A. That stays fairly stable. When I work to zero the offset and let cook for a while (say 15 minutes more) it may drift to 35mV. This is probably OK.
The trouble channel is I set for 1.75A and let cook. I walk back and it is is at 1.95A. I dial it down. Say 1.7. Then it drifts back up after 15 minute cook. I try to zero offset. After 15 minute cook 100mV or 150mv offset drift. repeat the cycle. It never settles to steady Iq or offset.
I have my 2nd set of boards and parts currently not assembled. I can remove a MOSFET from Kit #2 and replace it on trouble channel to see if that makes a difference. I can also do a Vgs check on those 2 loose MOSFETs to pick the best fit.
I thought I might try to switch SIT's between the channels. Or take the spare SIT's for the 2nd amp and switch one in. BUT if I assume Pras matching is good, my SITs are at 3.42V, 3.42V, 3.4V, 3.4V. They are all very close... So this is probably not the best approach?
Another thought was is build the other set of boards (I need to do it anyway), with very close attention to details. Measure each part, find any buggers mixed in, cross off each part on schematic when installing. Then visually compare the 2 boards to ensure all resistors are correct, diodes are correct, etc.
I can also go and measure voltages at certain points to see if that shows something wrong.
Another option, I do have a some matched quads of IRFP9240's on hand. Vgs for these are in range of 3.85 - not sure if that is OK, too high, to low? If I do use 9240's would other values on the SIT board need to change?
Thoughts?
My Sissysit #2 has some Iq creep as well, even after the heatsinks have reached temp. They drift by about 100mA and keep on doing so. That is why I have been testing them for 3 or more hours but they do eventually stop creeping. I have one channel connected to meters now and it starts around 1.2A Iq and after 3 hours, it stops climbing around 1.7A (or 170mV resistance). I will adjust it based on the 170 and then check it again after another entire heat cycle from cold. I am having some trouble with FE offset though and it jumps as well so I suspect it's the 100R pot and supposed to get new pots in today.
