Ho... Just install a speed potentio.. It don't really need fast speed something push the air is good enough
I am in the stage of thinking about the PSU of my USSA-5 build. I have already ordered the SLB boards. Now I need to figure out the transformer:
- Is +/- 24 DC rail voltage okay for the USSA-5? I am thinking of cranking the bias current up to 1.5A to 2A or so, depending on how hot the heatsinks get (I am happily running my Alephs at 1.9 A / 24 V rails in the same chassis).
- What AC voltage rating for the transformer secondaries would be right for 24 V rails with the SLB? I would guess that a Toroidy with 21 VAC would be about right, but it would be nice if someone could confirm / comment on this.
I got +/- 25.9Vdc on my SLB with a 600W, 22+22Vac trafo, drawing about 3.2A resistor load under test. SG main voltage is abit high at 235Vac.
So i think if you can get 21Vac trafo, it may be perfect.
For the PSU, the LT4320 gives about more than 2 Vdc more than standard diode bridge. I have not used the SLB but I believe that the voltage drop can be adjusted and reduced for less load current. In any case if it is about 2 or 2.5Vdc or so it is mostly compensated by the increase of the LT4320...
So in the end, chose your voltage transfo the same way as if it was a normal PSU. One volt difference is marginal. But as Harry3 indicates read post 880 regarding 4 or 8 speakers.
Fab
So in the end, chose your voltage transfo the same way as if it was a normal PSU. One volt difference is marginal. But as Harry3 indicates read post 880 regarding 4 or 8 speakers.
Fab
Something like this?
100W Ultimate Fidelity Amplifier
I do not know where it was posted originally, so quoting the Apex thread.
Dealing with the transition of class A to class B is one thing. Of course I like to keep it in class A for most of the time. My speakers are "amplifier friendly", so I am not too much concerned about the "4 Ohm vs 8 Ohm" issue (although I do understand this point). I also think that the "XYZ Ohm" rating method is almost useless in most situations, because it's usually more important if the impedance curve is smooth or has sharp peaks.
The other thing is that I want to bias the output devices such that they are operating in a more linear range. That's why I want to crank up the bias current. Looking at the Exicon curves, the linearity of the LatFETs is best between 2-6 A. That's why I thought to set the bias higher than suggested in the manual.
I like those test point pins. Where did you get them? Do you maybe have a Mouser part number?
mbrennwa,
Not sure, but after zooming looks like individual pins from break-away 2.54mm straight pin male header - something like Mouser part # 54102-T10-08 (example of an 8 pin header).
Another option used by diyAudio member Vunce for his build:
5002 Keystone Electronics | Mouser
Pics of Vunce's PCB:
USSA-5 Build with Review
Not sure, but after zooming looks like individual pins from break-away 2.54mm straight pin male header - something like Mouser part # 54102-T10-08 (example of an 8 pin header).
Another option used by diyAudio member Vunce for his build:
5002 Keystone Electronics | Mouser
Pics of Vunce's PCB:
USSA-5 Build with Review
Last edited:
I have used the same breakaway 2.54mm pins from the the long strip. Works perfectly fine and we can leave them on the boards after the setup as well. Can be helpful later and should be careful with the wiring not touching them during running any open wires. Otherwise test points are the best highlighted feature of Fab's boards, very very useful for the build process 
Thanks
Thanks
My USSA-5 will not have the power transistors and the NTC soldered to the PCB. I will use extension leads for the connection, because the PCB will not be on the heatsinks. I was wondering how to fix the NTC to the heatsink, and found this:
NTCALUG01A472H Vishay | Mouser Schweiz
Anything wrong with using this part instead of the one listed in the BOM/manual?
NTCALUG01A472H Vishay | Mouser Schweiz
Anything wrong with using this part instead of the one listed in the BOM/manual?
Hi
The BOM part has B parameter 3950K which is different from the ones you selected (3984K).
Electrical Isolation is absolutely required between chassis and thermistor.
As for extension leads with output transistors I am worried for possible oscillation.... at least gate resistor must be absolutely mounted on the transistor itself. Ensure to use an oscilloscope to check for oscillation.
Fab
Last edited:
I found this post. Hower, I couldn't find any problem reports reltated to this. Can you explain / point me to the problem?
That's a 1% difference. Also take a look at this comparison between the Vishay/lug NTC and the BOM NTC. I don't see any difference at all. What am I missing?
Yes, I am very well aware of this. I have done it in a few builds now, and I never ran into problems as long as the gate stopper was installed directly at the FET pins.
However, I just thought of something else that might be an issue in my situation. How important is thermal tacking between the power FETs? The reason I am asking is that I am planning to build my USSA-5 using an "Aleph" chassis, see here. There will be one power FET and one driver FET per heatsink to share the heat across the four sides of the chassis. The heatsinks will have some thermal contact with each other, but it's not like if the the power FETs would be mounted next to each other on the same heatsink. Also, the drivers would be distributed across the four heatsinks in the same way as the power FETs. So, here are a few questions related to this:
- How important is thermal tracking between the power FETs?
- How important is thermal tracking between the driver FETs?
- How important is thermal tracking between drivers and power FETs?
- What would be the best compromise for the pairing of the driver and power FETs on each heatsink? Positive power FET and positive driver on the same heatsink (and negative + negative on other heatsink)? Or positive power FET and negative driver on the same heatsink (and negative + positive on other heatsink)?
Last edited:
Hi
For the thermistor characteristics, I just pointed out that it was a different temperature curve . If you have checked that from 25C to about 65C it has only 1% difference than it should be ok.
“ How important is thermal tracking between the power FETs?” Important
“How important is thermal tracking between the driver FETs?“ Important
“How important is thermal tracking between drivers and power FETs?“ Important
“What would be the best compromise for the pairing of the driver and power FETs on each heatsink? Positive power FET and positive driver on the same heatsink (and negative + negative on other heatsink)? “Yes
“Or positive power FET and negative driver on the same heatsink (and negative + positive on other heatsink)?” no.
There are 2 thermistors, one for each polarity of the symmetrical circuit. Thus, as long as the heatsink location where is installed each thermistor for one channel (left or right) is the same then you should have the same temperature. Have in mind that P1 and P2 will be adjusted to center DC offset once the temperature has stabilized. However, temperature dissipation for positive and negative polarity should be the same. Some tweaking for that can also be done by moving one the thermistor closer to the output FET (a few millimeters for example) where is the max heatsink temperature. This way you can balance. It is easier with original part using the leads to move the thermistor on the heatsink location.
Drivers should be as close to the output transistors than it is meant if they were on the pcb.
As the distance on heatsink of each thermistor to the driver and output they should be the same as for the pcb.
As for the extension leads please ensure you have minimum length possible and I would sure check with a scope for the 10KHz square wave shape..
If you can show a sketch of your planned inside amplifier arrangements that would be great since it was never tried before for this amplifier. Thus you are really a FAB (Fearless Amplifier Builder)
Fab
For the thermistor characteristics, I just pointed out that it was a different temperature curve . If you have checked that from 25C to about 65C it has only 1% difference than it should be ok.
“ How important is thermal tracking between the power FETs?” Important
“How important is thermal tracking between the driver FETs?“ Important
“How important is thermal tracking between drivers and power FETs?“ Important
“What would be the best compromise for the pairing of the driver and power FETs on each heatsink? Positive power FET and positive driver on the same heatsink (and negative + negative on other heatsink)? “Yes
“Or positive power FET and negative driver on the same heatsink (and negative + positive on other heatsink)?” no.
There are 2 thermistors, one for each polarity of the symmetrical circuit. Thus, as long as the heatsink location where is installed each thermistor for one channel (left or right) is the same then you should have the same temperature. Have in mind that P1 and P2 will be adjusted to center DC offset once the temperature has stabilized. However, temperature dissipation for positive and negative polarity should be the same. Some tweaking for that can also be done by moving one the thermistor closer to the output FET (a few millimeters for example) where is the max heatsink temperature. This way you can balance. It is easier with original part using the leads to move the thermistor on the heatsink location.
Drivers should be as close to the output transistors than it is meant if they were on the pcb.
As the distance on heatsink of each thermistor to the driver and output they should be the same as for the pcb.
As for the extension leads please ensure you have minimum length possible and I would sure check with a scope for the 10KHz square wave shape..
If you can show a sketch of your planned inside amplifier arrangements that would be great since it was never tried before for this amplifier. Thus you are really a FAB (Fearless Amplifier Builder)
Fab
Last edited: