From what I understand , examining the speaker's impedance curve is the key - magnitude and phase angle .
If you look at the impedance curve for Troel's 3 way classic MK II , you'll see the magnitude of the impedance is 5 ohms and above ,
and the phase angle is well behaved . The amp almost sees a 5 ohm resistive load .
On paper , I figure the USSA - 3.2 would be a fabulous match to these speakers .
http://www.troelsgravesen.dk/SEAS-3-Way-Classic-mkII.htm
Recall, it takes current for a voice coil to build up a magnetic field .
So IMO its all about getting current from the power supply to the load ( speakers )
So a big enough power supply , and an amplifier with a low enough Zout is needed for this .
Usually , Zout is expressed as damping factor into an 8 ohm resistive load .
However , B&W's impedence curves are notorious for presenting a difficult load for an amp.
Its wild swings in the phase angle that give amplifiers grief .
I figure B&W 's where the minimum impedance 3.9 ohms , can be driven by amps with a DF >= 60 .
Like a H/K 3500 receiver can provide just enough current to drive these .
Where as B&W 's who's impedence curve drops down to 2.9 ohms - forget it . A beast of an amp is required .
I can't find the the impedance curve for the B&W 685 S2's . But the impedence drops down to 3.5 ohms .
So if a receiver can effectively drive these , a USSA 3.2 with a big power supply should be able to .
.
Fab ,
To add to this , I'm getting the components together to build a USSA 3.2 to drive a 1990's pair of B&W 603 s2's .
I wanted to build the USSA 3.2 , because F5 type of amps are so well behaved , and straightforward to get up an running .
For the B&W 603 s2's , the Impedence Curve drops down to 3.9 ohms .
I know an H/K 3400 receiver - 25 watts and with a DF = 60 can deliver just enough current to drive these speakers .
Can not find the impedence for the 603 s2's , but here is the impedence curve for the 603 s3's .... which drop down to 4.0 ohms .
https://www.stereophile.com/content/bw-dm603-s3-loudspeaker-measurements
The jFET's I bought from DiyAudio Stores have an Idss = 6 to 8 mA . So with the lower Idss , hopefully this will push up the Damping Factor a bit .
Following this 2 part article from Erno Borley , somewhere in here , he says to set Id to 60% to 70% of Idss
in order to maximize linearity .
https://audioxpress.com/article/JFETs-The-New-Frontier-Part-1
.
To add to this , I'm getting the components together to build a USSA 3.2 to drive a 1990's pair of B&W 603 s2's .
I wanted to build the USSA 3.2 , because F5 type of amps are so well behaved , and straightforward to get up an running .
For the B&W 603 s2's , the Impedence Curve drops down to 3.9 ohms .
I know an H/K 3400 receiver - 25 watts and with a DF = 60 can deliver just enough current to drive these speakers .
Can not find the impedence for the 603 s2's , but here is the impedence curve for the 603 s3's .... which drop down to 4.0 ohms .
https://www.stereophile.com/content/bw-dm603-s3-loudspeaker-measurements
The jFET's I bought from DiyAudio Stores have an Idss = 6 to 8 mA . So with the lower Idss , hopefully this will push up the Damping Factor a bit .
Following this 2 part article from Erno Borley , somewhere in here , he says to set Id to 60% to 70% of Idss
in order to maximize linearity .
https://audioxpress.com/article/JFETs-The-New-Frontier-Part-1
.
Right , I just checked on the single die Exicon ECX10n20 's , and the source is also connected to the metal contact area for the heat sink .
https://www.exicon.info/PDFs/ecx10n20.pdf
Double check this , but for the dual die Exicon ECW20n20's , the datasheet indicates there is no electronics connection to the metal contact area
for the heat sink .
https://www.exicon.info/PDFs/ecw20n20.pdf
.
Last edited:
" Amadeus00 said: " I tried to use UMS heatsinks in my build, but had to return them. UMS is not compatible with USSA-5 PCB layout. "
I misunderstood what was being said . The issue was that UMS heat sinks , sold by DIYAudio Stores , have pre drilled holes ,
that do not match the USSA boards .
https://diyaudiostore.com/products/ums-heatsinks
The issue is not an electronics conflict between p and n devices , for the USSA 3.2 and USSA 5 ( schematic as shown in post #4 ) ,
if they are electronically connected to the same heat sink .
Or have I just dug myself in even deeper ?
.
Yes, I did the drilling and tapping. It was the reason why I had to buy a bench drill 🤭. In my case … building this amp was something like economical suicide … but it was a lot of joy … support from FAB was great…. and it is still the best amp I have ever heard 🙂 And it is all that matters.
For USSA 5.0 REV 0.4 board , I'm measuring a mounting hole dia = 0.130 "
What mounting spacers or hardware did you use ?
I highly recommend buying an imperial Dial Gauge - It gets used all the time .
It can measure and be used to scribe a line into aluminum .
https://www.leevalley.com/en-ca/shop/tools/hand-tools/marking-and-measuring/calipers/32609-stainless-steel-dial-calipers?item=88N1701&utm_source=free_google_shopping&utm_medium=organic&utm_campaign=shopping_feed&utm_campaign=CAN-EN|PLA|Shopping|Hardware|Non-displayItems&gad_source=1&gclid=CjwKCAjwr7ayBhAPEiwA6EIGxL5FsjmmO_YUQtiN5pGTnG2VVadEaJA1Ne69WCPAUwonCsbkTgZnzRoCiyIQAvD_BwE
For a 4 - 40 machine screw
major dia = 0.112 " ( outside of the threads )
minor dia = 0.085 " ( inside of the threads )
M3 x 0.5 machine screw
major dia = 0.117 "
minor dia = 0.097 "
6 - 32 is too big
So it looks like a M3 x 0.5 machine screw is the best choice - but its only 0.005" larger in dia than a 4 - 40 .
I found these - which are threaded , so the machine screw goes right through .
https://www.mouser.ca/ProductDetail/Keystone-Electronics/24434?qs=UWqYQ/2cZWuTaOoOgBzi1g==
It would not be a haapy moment if the Tap broke in the heat sink , so suggest avoiding high carbon steel
and buying either HSS or tungsten .
For 6061 or 6063 aluminum , highly recommend using a cutting fluid - motor oil or baby oil is good .
.
Last edited:
I used one from TFF-M3X8/DR113BK and TFF-M3X5/DR113BK spacers. It was something like three years ago, so I cannot recall exactly which one of them I used. But the spacers are not big issue. At first, I just put the PCB without spacers on the heatsink and marked the mounting holes positions on the heatsink. Then again put the PCB on the heatsink again, but this time with spacers attached to the PCB. At this moment, you just try to align power transistors on the heatsink and decide, what length of spacers do you need. And also mark the holes for transistors holes on the heatsink as well.
In my case I decided to use M3 spacers, so for drilling, I used drill with 2,5mm diameter. Depth of the hole was something like 8mm, since the thickness of the heatsink was 10 mm and I did not want to go through the heatsink. It just does not look good. After your holes are drilled, you will need to use taping tool. But since I was not going through the heatsink, I needed to machine tap for blind-hole thread (M DIN 371, HSSE-Co 5, C Shape). You should use oil for drilling and taping, but the guy from mechanical shop once recommended me to use rapeseed oil. Some people use it also for lathe work as well.
In my case I decided to use M3 spacers, so for drilling, I used drill with 2,5mm diameter. Depth of the hole was something like 8mm, since the thickness of the heatsink was 10 mm and I did not want to go through the heatsink. It just does not look good. After your holes are drilled, you will need to use taping tool. But since I was not going through the heatsink, I needed to machine tap for blind-hole thread (M DIN 371, HSSE-Co 5, C Shape). You should use oil for drilling and taping, but the guy from mechanical shop once recommended me to use rapeseed oil. Some people use it also for lathe work as well.
manniraj said:
Yes I think my transformer is rated under VA compared to the required as per your recommendation.
But I think it should work fine with 252VA for both the channels.
Fab ,
Well , for the power supply for the USSA 3.2 , I was planning on using a single AnTek 300 VA toroid , with dual 18 Vac secondaries
... because I have one kicking around .
But .... as this amp is intended to drive my B&W 603 s2's , that drop down to 3.9 Ohms . it sounds like I should upgrade to a bigger transformer .
Its OK , money is no object .
Tell me 700 VA can provide enough muscle .
https://www.antekinc.com/au-7218-700va-18v-ul-transformer/
.
Yes I think my transformer is rated under VA compared to the required as per your recommendation.
But I think it should work fine with 252VA for both the channels.
Ok then you will see or....rather hear
Fab
Fab ,
Well , for the power supply for the USSA 3.2 , I was planning on using a single AnTek 300 VA toroid , with dual 18 Vac secondaries
... because I have one kicking around .
But .... as this amp is intended to drive my B&W 603 s2's , that drop down to 3.9 Ohms . it sounds like I should upgrade to a bigger transformer .
Its OK , money is no object .
Tell me 700 VA can provide enough muscle .
https://www.antekinc.com/au-7218-700va-18v-ul-transformer/
.
@Uunderhill,
I recently asked Antek about their different grade transformers, reply follows:
“AS series is for audio application with low noise, AN series is general application such as motor power supply, and AU series is industrial application with required UL.”
400VA is the biggest in the ‘AS’ model. You can go with two 300VA trafos in a dual mono power supply arrangement.
Or, order at Toroidy from Poland and get exactly what you want with a custom winding.
I recently asked Antek about their different grade transformers, reply follows:
“AS series is for audio application with low noise, AN series is general application such as motor power supply, and AU series is industrial application with required UL.”
400VA is the biggest in the ‘AS’ model. You can go with two 300VA trafos in a dual mono power supply arrangement.
Or, order at Toroidy from Poland and get exactly what you want with a custom winding.