Ixys Ixcp10m45s Ic
I have had these chips on back order from DigiKey since last year! I have received dozens of emails asking where to get them. None in stock anywhere. I got an email from DigiKey today stating that my back order has shipped. DigiKey has 5825 in stock as of this moment. Last time they had these in stock (November) they were gone in two weeks. If you need these, get some now.
I have had these chips on back order from DigiKey since last year! I have received dozens of emails asking where to get them. None in stock anywhere. I got an email from DigiKey today stating that my back order has shipped. DigiKey has 5825 in stock as of this moment. Last time they had these in stock (November) they were gone in two weeks. If you need these, get some now.
It's not just DigiKey. No one seems able to get these things. IXYS's own web site has a stock locator, which doesn't seem to show any stock at any of the major distributors for any of their useful (to tubeheads) high voltage chips or fets. I sent them at least 2 emails asking how to get small (100 to 500 pieces) of their products, and never got a reply. I guess they have enough big customers to be able to ignore everyone else.
The following article implies that the device is a complete chip:
http://ixdev.ixys.com/DataSheet/88ede06d-3a1c-4a5a-bab8-5b4e89b3323c.pdf
There was a paper on the IXYS web site explaining the use and application of this device. It also stated that the device is a complete circuit. That paper (called IXC.pdf) is no longer found on the IXYS web site, but can be found here:
http://www.datasheet4u.com/html/I/X/C/IXC_IXYSCorporation.pdf.html
I have seen other information that claims that the device is nothing more than a specially doped depletion mode mosfet. This article states that the device is a mosfet, and even implies that it can be replaced by a DN2540 by changing the resistor value. I have not tried this yet.
http://www.audioxpress.com/magsdirx/ax/addenda/media/jung2779.pdf
What's inside? I don't know. I have tried to cut the top off of the ones that I have freid, but always destroy the die in the process.
Drewaudio said:
You can request USPS shipping from Digikey. Prior to the recent USPS increases I was getting shipments from Digikey for around $25-40 USD via air mail. If you order, say, $100 worth of parts and compare Farnell and RS prices for similar you're still far ahead. Not worth it for a small order, but who couldn't use more Panasonic caps?
Well, they still show they got 5389 of them; I'm gonna get ten just because TL sez there's a problem. When I first learned of them, from Pete Millet, I ordered six not knowing there was any such availability problem; they showed hundreds in stock.
I'm glad to find I'm not the only one loves this chip. It's quiet as a grave, has a ridiculously high DR, and it's so much fun adjusting the draw in my stage with a screwdriver, just to hear what I can hear. We may have to buy smart, but at least they probably won't discontinue the thing.
Mahalo, TL,
Poinz
I'm glad to find I'm not the only one loves this chip. It's quiet as a grave, has a ridiculously high DR, and it's so much fun adjusting the draw in my stage with a screwdriver, just to hear what I can hear. We may have to buy smart, but at least they probably won't discontinue the thing.
Mahalo, TL,
Poinz
rdf said:
I think it's important to understand that Jung's approach is a way of calculating CCS impedance by measuring the rejection of an input voltage through a voltage divider, where the shunt leg is a one ohm resistor, and the series leg is the CCS under test. The resultant curves are not really noise curves per se (although the noise floor limits the measurements above a certain level of impedance depending on the frequency). They are 20 log of the inverse of the CCS impedance. Here's how Jung explained it in the first article of the series:
"IMPEDANCE OR NOISE REJECTION?
You should understand that current regulator circuits may be specified in terms of either equivalent dynamic impedance; i.e., “100kohm,” or rejection with respect to some applied voltage reference level. Here, the three calibration impedances equate to rejections of 80dB (10kohm), 100dB (100kohm), and 120dB (1Mohm). For impedances (Z) of more than 10kohm; (rejection of more than 80dB), you can use one of these approximations to convert between the two terminologies.
dB ~ 20 * log (1/Z) or Z ~ 10(dB/20)"
I applaud Jung's excellent tests, but I think he unnecessarily confused things by plotting the Y-axis in dB relative to one ohm (granted, this would be how the data came off the analyzer naturally). To me, it would have been more easily understood if he had inverted the Y-axis and plotted CCS impedances in decades as 10Kohm, 100kohm, 1Mohm, 10Mohm, ... etc., rather than as dB relative to one ohm. Still, the articles provide very valuable information. Interestingly to me, the lone LM317 does better than I would have guessed, not so different from a lone 10M45, but as he shows, cascoding it with something like a DN2540 makes it far better.
IXYS 10M45S (TO 220 package)
I have been wanting these babies for over a year now, so I went ahead and ordered 100 pcs.
I want to keep 20 or so. The rest can go to other diy'ers, so if you're interested send me a pm.
I will know the exact cost upon arrival, since there may be some customs charges and/or BTW (VAT) to cover as well.
I have been wanting these babies for over a year now, so I went ahead and ordered 100 pcs.
I want to keep 20 or so. The rest can go to other diy'ers, so if you're interested send me a pm.
I will know the exact cost upon arrival, since there may be some customs charges and/or BTW (VAT) to cover as well.
Angus, I think it depends on your application. I'm using them with ~70 volts across them, and about 10mA per, so about seven tenths of a watt. My major concern when I was building up the circuit was that I wasn't going to get the device (450 volts, 40 watts rating) 'on the cam' enough, and there would be noise or something.
In fact they work a treat. I put them in a slide-in heatsink, and now I'm laughing. I play the amp for 10 hours, turn it off, put my finger on the heatsink, and it's cooler than my finger!
Maybe one of these other savants who're using them for a plate load or something can give you a better read, but for me; I ain't gonna buy no more heatsinks, at least for this app. If you're burning five, ten watts, it's a plastic TO-220, so you probably better sink it, but if you're using it like me . . .
Aloha,
Poinz
In fact they work a treat. I put them in a slide-in heatsink, and now I'm laughing. I play the amp for 10 hours, turn it off, put my finger on the heatsink, and it's cooler than my finger!
Maybe one of these other savants who're using them for a plate load or something can give you a better read, but for me; I ain't gonna buy no more heatsinks, at least for this app. If you're burning five, ten watts, it's a plastic TO-220, so you probably better sink it, but if you're using it like me . . .
Aloha,
Poinz
As Poinz says, heat sinking really depends on your application. I have used these on the plates of driver tubes running off the same B+ as output tubes which requires them to drop a bunch of volts. In this case, since I was running the drivers at about 15mA and dropping close to 150V, I used a sink and had no issues.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.