Looks good! Shows clearly that we should not be pairing the E and D versions of the KSA1381/KSC3503.
The odd part is that your KSA1381E hfe seems low. I tested around 30~35 in the last few days, and they are all between 175-198, and 90% are between 180-195. (tested at 14~14.5 mA Ic).
Did you heatsink the TO-126, and let the temperature stabilize? If not, it would explain the difference.
Shall I send you a few to compare?
KSC3503E, J20/US, Fairchild
KSA1381E, B31/Korea, Fairchild
Project Status
First few sets of boards are shipping today. These are going mostly to a few people who have been following my posts in the PeeCeeBee thread for the last few months. If you are one of them, and depending on what you asked for, the package includes some or all of the following:
(2) Circuit Boards
(4) J112 jfets for the optional CCS
(2) Sets KSA1381E/KSC3503E, one per board
(4) Sets BC546C/BC556C (2 required, 2 spare) for operation over 45V
As of today, I also have a ship date for the first production order. I still have to check the FedEx delivery schedule, but that batch should be in my hands in about one week, and will start to ship around November 18.
First few sets of boards are shipping today. These are going mostly to a few people who have been following my posts in the PeeCeeBee thread for the last few months. If you are one of them, and depending on what you asked for, the package includes some or all of the following:
(2) Circuit Boards
(4) J112 jfets for the optional CCS
(2) Sets KSA1381E/KSC3503E, one per board
(4) Sets BC546C/BC556C (2 required, 2 spare) for operation over 45V
As of today, I also have a ship date for the first production order. I still have to check the FedEx delivery schedule, but that batch should be in my hands in about one week, and will start to ship around November 18.
Attachments
I have 100 each of the KSA1381E and KSC3503D. I did a quick DMM sort on all when I got them. Of those 93 KSA were in the 140's and the other 7 were in the 150's; of the KSC 52 were in the 70's, 46 were in the 80's and 2 were in the 90's. The codes are C21 for the KSC and C29 for the KSA and IIRC are of Korean manufacture.
They were not heat sinked. The test is so fast that unless you use large amounts of averaging there is no discernible heat is generated; they basically remain at room temperature. This is the likely cause of the lower figures. We would get a little added gain in practice due to the higher temperatures they would be operating at.
As far a some spares, if you can afford to send a couple extra to see how a small sample compare, I'd be happy to do so. Would you like any of my KSA1220AY / KSC2690AY devices cross-shipped for your own investigation? If so, you have my email and can communicate addresses and such off board.
They were not heat sinked. The test is so fast that unless you use large amounts of averaging there is no discernible heat is generated; they basically remain at room temperature. This is the likely cause of the lower figures. We would get a little added gain in practice due to the higher temperatures they would be operating at.
As far a some spares, if you can afford to send a couple extra to see how a small sample compare, I'd be happy to do so. Would you like any of my KSA1220AY / KSC2690AY devices cross-shipped for your own investigation? If so, you have my email and can communicate addresses and such off board.
Pete,
are you planning on posting the Bottom Copper trace for those that wish to Etch their own.? Post #65 has everything Except the bottom copper.
are you planning on posting the Bottom Copper trace for those that wish to Etch their own.? Post #65 has everything Except the bottom copper.
C29 was one of the two date codes which tested lower, and was hard to match (according to my notes, from Mouser). So, who knows.
The KSA1381-ES, B31/Korea, was the one that tested very close to the NOS KSC3503-E. Almost all could be sorted into pairs that were less than 10% apart at 14mA, and 40% pairs that were less than 5%.
I will send you one of the sets that were sorted by hfe, and a few unsorted parts. It would be most helpful if you could verify what I measured manually with your curve tracer. I will send a PM later today.
Yes, sorry. Not reviewed yet, use at your own discretion, and if possible post comments for other people who may want to use it.
OK. That means you have one of the date codes which tests a bit on the low hfe side.
Most of mine tested around 190. Maybe (I hope) Jason can shed more light on this in a few days... 😕
Hi PMI
Looking at the schematics again I believe that the biasing scheme with the 2 diodes will not always work since the 2sk1058/J162 can have each a VGSoff as high as 1.45v . If the sum of VGSoff of both lateral is higher than 1.3v than the diodes should not be populated to avoid clamping the voltage, so the bias is only done with R19. Can you confirm.
Thanks
Fab
Looking at the schematics again I believe that the biasing scheme with the 2 diodes will not always work since the 2sk1058/J162 can have each a VGSoff as high as 1.45v . If the sum of VGSoff of both lateral is higher than 1.3v than the diodes should not be populated to avoid clamping the voltage, so the bias is only done with R19. Can you confirm.
Thanks
Fab
Good question. Perhaps this relies on the supply voltage being a fraction of the specified Vds maximum.
I took that as accepted from the PeeCeeBee thread, and that is what I have been using for three revisions of the board. Does not make it right, of course.
The range for that parameter, from min to max is very large, 0.15 to 1.5V. Logically, the minimum would apply at very low Vds. I am not sure one can conclude that the opposite is true, but perhaps using the two diodes relies on the supply voltage being relatively low, compared to the maximum for the part.
I continue to add names to the list. If it keeps growing, I will open a GB thread.
$20/set of 2 boards, including the four jfets for the CCS, and not including the other parts in the pic, which you may or may not want. Shipping and Paypal adds about 3.75 in the US. International shipping, IF I can get away with using regular mail $5.50 (First Class Package International, adds about $14). I will update some details later this weekend.
The first people get boards from the first board orders, which seems fair to me... also:
To be honest, the other advantage (for me) is that I do not have to deal with too many payments, addresses, or packages at one time. The US Post Office is not making International shipments easy these days... 🙄
I'd like 2sets w/CCS fets. I'm in MKE so can pick up and pay in person or use PP, your choice. You have email from me.
Thanks.
Thanks.
Added dady and dshortt9 to the list.
The Jfets are included with all boards.
They only cost 10 cents in quantity 100 when I was buying them for the Cap Multiplier boards, but between 50 cents and $1 each. Some people can't get them easily, or have to pay separate postage, etc... 'nuf said.
Hey Pete,
I think you should have me down for 6 boards (I hope). With all parts you can provide if possible..
-Raja
I think you should have me down for 6 boards (I hope). With all parts you can provide if possible..
-Raja
Any advantage/disadvantage to using the BC546C/556C in place of the BC550/560? Also, what components need to be matched?
Thanks, Terry
Thanks, Terry