And make sure you use the V2 schematics posted here :
UDNeSS, or You don't need Semisouth's
It is implicit that you have a pair of electrolytic caps close to the amp PCB.
For example 10,000µF 35V.
LKG1V103MESBBK Nichicon | Capacitors | DigiKey
LKG1V103MESBBK Nichicon | Mouser Europe
Patrick
UDNeSS, or You don't need Semisouth's
It is implicit that you have a pair of electrolytic caps close to the amp PCB.
For example 10,000µF 35V.
LKG1V103MESBBK Nichicon | Capacitors | DigiKey
LKG1V103MESBBK Nichicon | Mouser Europe
Patrick
Yes - I have the updated pcbs, thanks for verifying.
When you say capacitors need to be close - how close? My PSU board will be less than 6-inches away. Is that close enough?
> I have the updated pcbs
Then why did you ask to match IRFP9240's ?
> how close?
FirstWatt J2
Patrick
Then why did you ask to match IRFP9240's ?
> how close?
FirstWatt J2
Patrick
Post #126,127.
> How would I determine accurate current on J113 ?
Multimeter on a breadboard.
Or meaure current with DMM on the UDNeSS PCB before soldering the MOSFETs.
Just roughly 6mA to start with.
You need to trim later for DC offset anyway,
Rdg means degeneration resistor, not drain to gate.
Patrick
> How would I determine accurate current on J113 ?
Multimeter on a breadboard.
Or meaure current with DMM on the UDNeSS PCB before soldering the MOSFETs.
Just roughly 6mA to start with.
You need to trim later for DC offset anyway,
Rdg means degeneration resistor, not drain to gate.
Patrick
Thanks for clarifying for me.
hopefully one last question before I place my order.
Your schematic shows the Servo (C101, C102) as 1uF, but your BOM list and part# indicates a 0.22uF Capacitor?
Can you tell me which is correct? 1uF or 0.22uF, 50Vdc 1206 SMD part?
I have triple checked the Schematics/BOMs and made a couple last minute part substitutes due to inventory, but I think I am very close.
> 1uF or 0.22uF, 50Vdc 1206 SMD part?
The PCB will take 1206 devices.
It needs to be minimum 25V.
I want to use a better dielectric material, namely C0G.
That is only available up to 0.22uF.
The servo will then have corner frequency of 0.7Hz, which is sufficient.
But the part is still on order at Mouser.
So perhaps you want to try without servo first.
Alpha tester's experience is that servo is not necessary.
But that assumes you have sufficient heatsink.
So test first before deciding.
Patrick
The PCB will take 1206 devices.
It needs to be minimum 25V.
I want to use a better dielectric material, namely C0G.
That is only available up to 0.22uF.
The servo will then have corner frequency of 0.7Hz, which is sufficient.
But the part is still on order at Mouser.
So perhaps you want to try without servo first.
Alpha tester's experience is that servo is not necessary.
But that assumes you have sufficient heatsink.
So test first before deciding.
Patrick
2sk170 BL grade for Udness?
Hi Patrick,
The other Patrick, from the Great State of Minasooota, is graciously sending me a set of boards for this Super-Cool project of yours! 2 questions.
1. Are BL grade Toshiba's appropriate for this bulid? And how to figure Q1 if so?
2. Is a 4U, "jack of all cases," enough heat sink to run without needing the servo? I have an Aleph J that I have been wanting to mod but would love to slip "Udeness" right in the chassis while I tinker, heck thats why i bought the chassis in the first place.
Thank You,
Jeffrey
Hi Patrick,
The other Patrick, from the Great State of Minasooota, is graciously sending me a set of boards for this Super-Cool project of yours! 2 questions.
1. Are BL grade Toshiba's appropriate for this bulid? And how to figure Q1 if so?
2. Is a 4U, "jack of all cases," enough heat sink to run without needing the servo? I have an Aleph J that I have been wanting to mod but would love to slip "Udeness" right in the chassis while I tinker, heck thats why i bought the chassis in the first place.
Thank You,
Jeffrey
How long does it take to stabilize for matching the jfets?
I am starting with the j113s figuring they are a little robust, but at 3.5mA and 9Vdc, I get about 237mV, but steadily goes down to 233, before ticking back to 237 and starting back down again.
Fearing I did something wrong, I stopped after a minute, then tried at 5.8mA and get same thing but starts at 496mV and goes down to 483mV before starting over back at 496mV.
I only let it go for 45-60 secs since it was a higher current.
Does this sound right? When should I notate the Source voltage for matching them? Should I just set a timer and document the voltage at say 2 mins?
I am starting with the j113s figuring they are a little robust, but at 3.5mA and 9Vdc, I get about 237mV, but steadily goes down to 233, before ticking back to 237 and starting back down again.
Fearing I did something wrong, I stopped after a minute, then tried at 5.8mA and get same thing but starts at 496mV and goes down to 483mV before starting over back at 496mV.
I only let it go for 45-60 secs since it was a higher current.
Does this sound right? When should I notate the Source voltage for matching them? Should I just set a timer and document the voltage at say 2 mins?
Breathing over a JFET will also change the Idss / Vgs. 
For the J113 used as current source, exact value is not important, as you can trim the working point.
For the 2SK209GR used in the input LTP, the SOT23 test socket will give you some stabilisation.
But as Morde said, 2% is fine.
It is not a fully symmetrical, low distortion amplifier like the F5X.
So matching is not so critical.
Patrick
For the J113 used as current source, exact value is not important, as you can trim the working point.
For the 2SK209GR used in the input LTP, the SOT23 test socket will give you some stabilisation.
But as Morde said, 2% is fine.
It is not a fully symmetrical, low distortion amplifier like the F5X.
So matching is not so critical.
Patrick
Thanks for reply’s makes me feel better that I shouldn’t blow any of these up during testing. Interesting comparison, as I tested with a Chinese tester from eBay and then with your matching jig. Not done yet, but was close on the FQAs and been fairly close on the j113s after about half have been tested.
I might have an issue - so I want to check with the pro’s.
Testing the 2SK209s at 2.1mA and measuring the Vsource, I am only registering 0.3-0.4mV on my meter? I have quadruple checked and even flipped the voltage and tested a FQA at 1.2A and seem to get fairly consistent at 4.3v.
Do you think I have the jig set up incorrect? I read the data sheet and says at VDS=10v (I have the PSU at 10Vdc) then VGS=0, which seems to validate my testing, but want to be sure.
I tested 5 so far and all were from 0.3-0.5mV
Testing the 2SK209s at 2.1mA and measuring the Vsource, I am only registering 0.3-0.4mV on my meter? I have quadruple checked and even flipped the voltage and tested a FQA at 1.2A and seem to get fairly consistent at 4.3v.
Do you think I have the jig set up incorrect? I read the data sheet and says at VDS=10v (I have the PSU at 10Vdc) then VGS=0, which seems to validate my testing, but want to be sure.
I tested 5 so far and all were from 0.3-0.5mV