Rank: Member
Groups: Member
Joined: 11/20/2010(UTC)
Posts: 17
Location: Shanghai China
|
Hello Russ, Brian and DIYers, I have started building my Placid, and I just noticed I forgot to order the PN4393 from Mouser.....  Shipping charges are quite high where I live and I'd hate to place an order just for that part... Since it is just (I assume) a current source, I was wondering if the part is critical or if another model/type of J-FET would be suitable as a replacement? What I have on hand are Junction Types (2SK117, 2SK363, BF245B) and "Enhancement Mode" (BS170) Alternatively (you may not want to look at all those datasheets....), would anyone mind advising what is the current required through the LEDs ? Thanks ! Edited by user Friday, March 18, 2011 9:28:19 PM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 11/20/2010(UTC)
Posts: 17
Location: Shanghai China
|
Edit+Edit 09/04/2011 Well I thought I found the solution, trying a 2SK363GR in place of the PN4393. It did seem to work ... but I can't get the shunt current high enough with that part..... seems I 'd have to play with the resistor in the current source, and the solution is not elegant since I have to twist the pins to fit the part.... Nevermind, I'll buy the part..... just will cost me 15 US$... Edited by user Saturday, April 9, 2011 4:26:56 AM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 11/20/2010(UTC)
Posts: 17
Location: Shanghai China
|
Haha some news !! I got it to work with the 2SK363 and fiddling with the resistor. Problem, it requires a quite low value now to produce enough current to light up the LEDS. Problem is.. the shunt current is drifting as the whole thing gets hot. (and drifting dangerously.... it went up to 365mA from 350mA after a few minutes.... didn't want to go further so stopped it) Did anyone experience the same with the original PN4393 ? Now I'm thinking of tweaking my Placid to be closer to a PlacidBP, which looks like a more robust solution .... (just 2 PCB traces to cut and here I go.. first Placid 2.1.2 in History !) Edited by user Tuesday, April 12, 2011 7:28:26 AM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 11/20/2010(UTC)
Posts: 17
Location: Shanghai China
|
OK, let's continue this one man thread... I modified my Placid to act like a Placid BP 2.1.2, that is to say with a LED setting the voltage on QN1 instead of a current source. The shunt current continues to drift, so finally it was not coming from the JFET. Suddently came to my mind another observation I made a while ago: the potentiometers that set the shunt current are not designed to stand the poser through them. Just make the math: 350 mA through 2.8ohms (what I am measuring) = 0.343 W These little pots are designed to dissipate 0.5W..... but that's on 100% of the resistive element surface. Now we're using a fraction of it.. no doubt the thing is getting hot ! Edit: Measurements using precision ohm-meter Edited by user Friday, April 15, 2011 10:28:51 AM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 11/20/2010(UTC)
Posts: 17
Location: Shanghai China
|
Some more proof of my "theory" When cold, the potentiometer reads 2.8 ohm .... when hot (after 5 minutes), it's 1.7 ohm It's a 200 Ohm , as per Twisted Pear BOM. Now I remember... the Placid BP 2.1.2 has a 20 Ohm Pot in the BOM (unlike the schematic, which is wrong)... So we're using just about 5% of a surface that's supposed to dissipate half a Watt... so I desoldered the thing , which is a pain in the butt being a 3 legs one BTW,.... and soldered a 20ohm instead RESULT .... The shunt current is still drifting, ] (*,) but much slower than before. It's drifting at a rate of ~1mA per minute (was about ~10mA per minute) Again, the resistance of the potentiometer is decreasing when hot (which by the way surprises me as I thought it would have been the contrary...) Anyone has an idea of the temperature coefficient of those potentiometers ? (not sure what is the technology of these mini pots..) Edited by user Friday, April 15, 2011 10:29:12 AM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Lazybutt wrote:Some more proof of my "theory"
When cold, the potentiometer reads 2.8 ohm .... when hot (after 5 minutes), it's 1.7 ohm
It's a 200 Ohm , as per Twisted Pear BOM. Now I remember... the Placid BP 2.1.2 has a 20 Ohm Pot in the BOM (unlike the schematic, which is wrong)...
So we're using just about 5% of a surface that's supposed to dissipate half a Watt... so I desoldered the thing , which is a pain in the butt being a 3 legs one BTW,.... and soldered a 20ohm instead
RESULT
....
The shunt current is still drifting, ] (*,) but much slower than before.
It's drifting at a rate of ~1mA per minute (was about ~10mA per minute)
Again, the resistance of the potentiometer is decreasing when hot (which by the way surprises me as I thought it would have been the contrary...)
Anyone has an idea of the temperature coefficient of those potentiometers ? (not sure what is the technology of these mini pots..)
That is very strange. My pots show no such change when warm. :) Neither 200R or 20R are wrong. :) Use 10-20R. Also do not forget the 1-2R in series. 20R is better than 200R for most cases because it utilized a broader range of the pot for most apps. that is why we supply it. The reason you may be seeing changes in voltage or current is not tempco of the pot, but rather of the LEDs and the transistors themselves. This can be controlled with adequate heat sinking and ventilation. The supply current and voltage will stabilize with the temperature. Remember the output voltage will start out higher(cold) than it ends up(warm). So always adjust for the highest case. Which is when the supply is cold. Edited by user Friday, April 15, 2011 10:40:45 AM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
The reason we supply a schematic is so the people who want to push it hard have all they need to know to adjust parts such as using lower value CCS pots. :) Thats why it is DIY use only.
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 11/20/2010(UTC)
Posts: 17
Location: Shanghai China
|
Thanks for replying Russ ! Yes the Tempco of the LEDs have an effect, but in my case, after changing ONLY VR1, it had an effect on the drifting rate. So definitely the pot has an effect. Now I may have an explanation: The Pots you are using are probably Cermet, which have a positive tempco. Resistance tends to increase when hot, which tends to reduce the shunt current => the system is stable My pots are carbon (just broke one apart to check...  ) I admit buying them off eBay.. seems they are fake Bourns..) yes they were cheap... 6 euro cents a piece Carbon has a negative tempco: when hot, the resistance decreases, which increases the shunt current, which increases heat, which decreases resistance, etc..... => the system is unstable because I use them at the limit of their capacity (after accurate calculations, 2.8 ohm @ 350 mA is equivalent to 0.1 Watt, but thats only on 14% of the resistive element surface, so "equivalent" to 0.7 Watt on all surface ..which is why they're getting hot) At 250 mA, the Placid is not drifting I'll probably just go for a fixed resistor PS: what I learned.. only buy parts from verified suppliers, you may save money, but then you're putting at risk a more expemnsive setup... a Buffalo in my case!
|
|
|
|
|
|
Forum Jump
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.
