On Tuesday, 30 October 2018 01:34:26 UTC, jjhu...@gmail.com wrote:> On Monday, October 29, 2018 at 6:03:58 PM UTC-4, John Larkin wrote: > > On Mon, 29 Oct 2018 14:57:20 -0700, John Larkin > > <jjlarkin@highland_snip_technology.com> wrote: > > >On Mon, 29 Oct 2018 14:39:08 -0700 (PDT), three_jeeps > > > wrote:> > >If power doesn't matter, there are generic gumdrop NPNs and PNPs and > > >they are all pretty much the same. Casettes are noisy all by > > >themselves, so the transistors won't matter. > > > > > >I think there is a reason that people like Tek used to sell curve > > >tracers and transistor testers, and don't any more. > > > > > >I'd like a box that characterizes RF fets (phemts, SiC, GaN) for both > > >DC specs and capacitance, but we do that other ways now, and wouldn't > > >pay a lot for a dedicated box. > > > > Keysight has a curve tracer, the B1505A Power Device Analyzer. It's > > around $80K. > > > > We have a couple of their SMUs, which can be programmed to do curve > > tracing, and lots more.> Thank you. A bit pricey for my needs but if they come down to <= $200 USD, I'd consider it. > JI can't help thinking that with the help of a bit of eyesight persistence one could very simply get a scope to display a set of curves without even so much as resorting to a microprocessor. The mu-p would make it easier I expect, but the point is it doesn't look like a tricky problem. NT
pointers to transistor (semiconductor) curve tracer circuits/projects
Started by ●October 28, 2018
Reply by ●October 29, 20182018-10-29
Reply by ●October 29, 20182018-10-29
On Monday, October 29, 2018 at 9:39:33 PM UTC-4, tabb...@gmail.com wrote:> On Tuesday, 30 October 2018 01:34:26 UTC, jjhu...@gmail.com wrote: > > On Monday, October 29, 2018 at 6:03:58 PM UTC-4, John Larkin wrote: > > > On Mon, 29 Oct 2018 14:57:20 -0700, John Larkin > > > <jjlarkin@highland_snip_technology.com> wrote: > > > >On Mon, 29 Oct 2018 14:39:08 -0700 (PDT), three_jeeps > > > > wrote: > > > > >If power doesn't matter, there are generic gumdrop NPNs and PNPs and > > > >they are all pretty much the same. Casettes are noisy all by > > > >themselves, so the transistors won't matter. > > > > > > > >I think there is a reason that people like Tek used to sell curve > > > >tracers and transistor testers, and don't any more. > > > > > > > >I'd like a box that characterizes RF fets (phemts, SiC, GaN) for both > > > >DC specs and capacitance, but we do that other ways now, and wouldn't > > > >pay a lot for a dedicated box. > > > > > > Keysight has a curve tracer, the B1505A Power Device Analyzer. It's > > > around $80K. > > > > > > We have a couple of their SMUs, which can be programmed to do curve > > > tracing, and lots more. > > > Thank you. A bit pricey for my needs but if they come down to <= $200 USD, I'd consider it. > > J > > I can't help thinking that with the help of a bit of eyesight persistence one could very simply get a scope to display a set of curves without even so much as resorting to a microprocessor. The mu-p would make it easier I expect, but the point is it doesn't look like a tricky problem. > > > NTYou are probably correct in that regard, for the need I have. I am considering longer range applications where I'd like numbers to draw curves from. I tend to over-engineer solutions sometime. OTOH, I think it would be a neat gadget to have an experiment with. I might learn something useful that I didn't know, since this isn't my daily engineering work. Thanks J
Reply by ●October 29, 20182018-10-29
three_jeeps wrote:> > Phil Allison wrote: > > > > > > I have found a few transistors that are dead and need to be replaced. > > > Fortunately, they are in parts of the amp where the other channel has > > > an identic transistor. > > > Having one transistor as a base will allow me to match something close > > > to the original. > > > > > > I have made (in the past) the typical transistor curve tracer that shows > > > the curves on a scope - cute but not precise enough for my intentions. > > > I am more interested in a lab type instrument. > > > > > > > ** You are wasting your time, that is no way to even get close. > > > > I have a LOT of experience with substituting transistors found in > > all brands of hi-fi and pro audio electronics. There are plenty of > > clues to be found & used in choosing suitable replacements for obsolete > > and code numbered parts. > > > > Sometimes devices need to be Vbe and/or Hfe matched - circuit topology > > is the big clue there. > > > > Hint: watch out for TO3 devices that seem like ordinary power > > transistors but are in fact Darlingtons. > > > > > > I agree with you about the circuit clues. Since my background is > math/control theory and hardware/digital architect (as in design of CPUs), > my analog skills have gotten a bit rusty. >** Assuming that using a curve tracer is the answer to your prayers is worse than rusty.> If I have some questions, perhaps I can bounce them off you privately ? >** Better post them right here. However, without actual units to do tests on and/or clear schematics, the task can become near impossible. I can point out what clues to look for but you have to find and make good use of them. Experience with actual repairs soon revealed to me that manufacturers mostly chose particular semis ( BJTs, JFETs etc) because they were cheap and readily available - or at least were to them at the time. Not because their electrical characteristics were so special nothing else would do as well or better. Having said that, one does need to get the basics right - ie substitutes should be in the same package, have sufficient breakdown voltage, power handling, Hfe, correct polarity etc. If devices need matching ( ie input diff pairs or output BJTs in parallel ) - then knowing how to do that job reliably is important too. .... Phil
Reply by ●October 29, 20182018-10-29
On Tuesday, 30 October 2018 01:50:35 UTC, jjhu...@gmail.com wrote:> On Monday, October 29, 2018 at 9:39:33 PM UTC-4, tabby wrote: > > On Tuesday, 30 October 2018 01:34:26 UTC, jjhu...@gmail.com wrote: > > > On Monday, October 29, 2018 at 6:03:58 PM UTC-4, John Larkin wrote: > > > > On Mon, 29 Oct 2018 14:57:20 -0700, John Larkin > > > > <jjlarkin@highland_snip_technology.com> wrote: > > > > >On Mon, 29 Oct 2018 14:39:08 -0700 (PDT), three_jeeps > > > > > wrote: > > > > > > >If power doesn't matter, there are generic gumdrop NPNs and PNPs and > > > > >they are all pretty much the same. Casettes are noisy all by > > > > >themselves, so the transistors won't matter. > > > > > > > > > >I think there is a reason that people like Tek used to sell curve > > > > >tracers and transistor testers, and don't any more. > > > > > > > > > >I'd like a box that characterizes RF fets (phemts, SiC, GaN) for both > > > > >DC specs and capacitance, but we do that other ways now, and wouldn't > > > > >pay a lot for a dedicated box. > > > > > > > > Keysight has a curve tracer, the B1505A Power Device Analyzer. It's > > > > around $80K. > > > > > > > > We have a couple of their SMUs, which can be programmed to do curve > > > > tracing, and lots more. > > > > > Thank you. A bit pricey for my needs but if they come down to <= $200 USD, I'd consider it. > > > J > > > > I can't help thinking that with the help of a bit of eyesight persistence one could very simply get a scope to display a set of curves without even so much as resorting to a microprocessor. The mu-p would make it easier I expect, but the point is it doesn't look like a tricky problem. > > > > > > NT > > You are probably correct in that regard, for the need I have. I am considering longer range applications where I'd like numbers to draw curves from. I tend to over-engineer solutions sometime. > OTOH, I think it would be a neat gadget to have an experiment with. I might learn something useful that I didn't know, since this isn't my daily engineering work. > Thanks > JYou could always add little glitches on the display every 10 units or so. But as Phil more or less said, modern commercial circuits are designed to not care about a fair spread of transistor values. NT
Reply by ●October 30, 20182018-10-30
On Mon, 29 Oct 2018 18:25:01 -0700 (PDT), jjhudak4@gmail.com wrote:>On Monday, October 29, 2018 at 8:43:04 PM UTC-4, John Larkin wrote: >> On Mon, 29 Oct 2018 23:30:36 +0100, Gerhard Hoffmann >> wrote: >> >> >Am 29.10.18 um 22:57 schrieb John Larkin: >> >> On Mon, 29 Oct 2018 14:39:08 -0700 (PDT), three_jeeps >> >> >> >>> @bitrex >> >>> You are either a troll or had way too much coffee. GIve it a rest man.I have some high end audio gear from the 80s and I do some repair for friends/clients. I also do format conversions for myself and other ppl. >> >>> I posted a question and 60% of the drivel that flowed is totally unrelated or begged the question about usefulness. I always thought that an IQ test should be given to ppl before they got an on-ramp to the net. Most would fail the test. >> >>> I don't care about your technology views, your likes/dislikes, or your perspective on audio or electronics, your political view, or what your father did or did not do. Mine was a WWII vet - fighter pilot but has nothing to do with the question I asked. >> >>> >> >>> Why a cassette deck? I am repairing one for a friend. He and I both have a Nakamichi 1000ZXL. It appears his preamp died. I didn't challenge him as to why he *still* uses it...to each their own. >> >>> take a back seat, shut up, and enjoy the ride, or leave. >> >> >> >> You sound obnoxious. This is an unmoderated newsgroup, and we discuss >> >> electronics. If you want paid consultants, you'll have to pay for >> >> them. >> >> >> >> If power doesn't matter, there are generic gumdrop NPNs and PNPs and >> >> they are all pretty much the same. Casettes are noisy all by >> >> themselves, so the transistors won't matter. >> >> >> >> I think there is a reason that people like Tek used to sell curve >> >> tracers and transistor testers, and don't any more. >> >> >> >> I'd like a box that characterizes RF fets (phemts, SiC, GaN) for both >> >> DC specs and capacitance, but we do that other ways now, and wouldn't >> >> pay a lot for a dedicated box. >> > >> >He asked an on-topic question and got shitty, off-topic answers, >> >all of them but maybe 2 or 3. >> >Your's was none of the 3. >> > >> >Gerhard >> >> I answered first, and my response was civil and relevant. He didn't >> answer my question about schematics. >> >> He apparently resents not getting free consulting, and resents >> discussion drift. It's an unmoderated public forum; get used to it. >> >> >> -- >> >> John Larkin Highland Technology, Inc >> picosecond timing precision measurement >> >> jlarkin att highlandtechnology dott com >> http://www.highlandtechnology.com > >I did reply to your first post but I forgot and did it in my email and it didn't get posted. To address your points: >1) I usually try to find schematics, sometimes not very successful. If I do have schematics, I try to model it to see how it performs (if it is something I am not familiar with). I did answer the question of noise. >2) There are lots of reasons why curve tracers are no longer made by Tek, HP, etc. My speculation is that as a business decision, it is not profitable because ppl solve the design problem with different technology rather than with discrete transistors. That does not imply the use/application is useless.Small-signal bipolar transistors just aren't very interesting; all the curves look about the same. Beta is all over the place anyhow. For audio, just get a decent part with good beta and enough Vcbx or something. A curve tracer won't tell you Ft or capacitances or noise. A curve traces is more useful for fets, especially RF parts where the data sheet often has no DC data.>3) I know a curve tracer does not provide any insight as to noise characteristic, and I know that noise is a concern - especially in old clunky electromechanical beasts like cassette deck. If you look at some of the designs and implementations, some of the low noise solutions are quite elegant. >4) Discussion drift - that is true, I resent it. I ask a technical question on a technical group and expect the answers to be technical. As one reader pointed out, most of it was not. >5) free consulting - is appreciated, and thanked in kind. Just because an approach doesn't do something doesn't mean the approach is invalid for other things. Because curve tracers do not provide insight on noise characteristics doesn't me they aren't useful for their intended purpose. I have a fairly solid academic background in EE but admittedly rusty in certain areas. I took extra courses in semiconductor physics even though it was not my major and that was many years ago. > >6) unmoderated public forum. Right - anybody can say anything on any topic. My expectations for ppl to stay on topic was apparently incorrect.It wouldn't be very interesting if everyone stayed on topic. This is a discussion group. This is sci.electronics.DESIGN. Design implies a lot of meandering. -- John Larkin Highland Technology, Inc lunatic fringe electronics
Reply by ●October 30, 20182018-10-30
On 10/29/2018 05:39 PM, three_jeeps wrote:> @bitrex > You are either a troll or had way too much coffee. GIve it a rest man.I have some high end audio gear from the 80s and I do some repair for friends/clients. I also do format conversions for myself and other ppl. > I posted a question and 60% of the drivel that flowed is totally unrelated or begged the question about usefulness. I always thought that an IQ test should be given to ppl before they got an on-ramp to the net. Most would fail the test. > I don't care about your technology views, your likes/dislikes, or your perspective on audio or electronics, your political view, or what your father did or did not do. Mine was a WWII vet - fighter pilot but has nothing to do with the question I asked. > > Why a cassette deck? I am repairing one for a friend. He and I both have a Nakamichi 1000ZXL. It appears his preamp died. I didn't challenge him as to why he *still* uses it...to each their own. > take a back seat, shut up, and enjoy the ride, or leave. >learn how Usenet works, this isn't your Facebook wall, or Twitter (what's this "@" stuff?) I responded once to your original post the other replies were to other people.
Reply by ●October 30, 20182018-10-30
On 10/29/2018 08:42 PM, John Larkin wrote:> On Mon, 29 Oct 2018 23:30:36 +0100, Gerhard Hoffmann > <ghf@hoffmann-hochfrequenz.de> wrote: > >> Am 29.10.18 um 22:57 schrieb John Larkin: >>> On Mon, 29 Oct 2018 14:39:08 -0700 (PDT), three_jeeps >>> <jjhudak@gmail.com> wrote: >>> >>>> @bitrex >>>> You are either a troll or had way too much coffee. GIve it a rest man.I have some high end audio gear from the 80s and I do some repair for friends/clients. I also do format conversions for myself and other ppl. >>>> I posted a question and 60% of the drivel that flowed is totally unrelated or begged the question about usefulness. I always thought that an IQ test should be given to ppl before they got an on-ramp to the net. Most would fail the test. >>>> I don't care about your technology views, your likes/dislikes, or your perspective on audio or electronics, your political view, or what your father did or did not do. Mine was a WWII vet - fighter pilot but has nothing to do with the question I asked. >>>> >>>> Why a cassette deck? I am repairing one for a friend. He and I both have a Nakamichi 1000ZXL. It appears his preamp died. I didn't challenge him as to why he *still* uses it...to each their own. >>>> take a back seat, shut up, and enjoy the ride, or leave. >>> >>> You sound obnoxious. This is an unmoderated newsgroup, and we discuss >>> electronics. If you want paid consultants, you'll have to pay for >>> them. >>> >>> If power doesn't matter, there are generic gumdrop NPNs and PNPs and >>> they are all pretty much the same. Casettes are noisy all by >>> themselves, so the transistors won't matter. >>> >>> I think there is a reason that people like Tek used to sell curve >>> tracers and transistor testers, and don't any more. >>> >>> I'd like a box that characterizes RF fets (phemts, SiC, GaN) for both >>> DC specs and capacitance, but we do that other ways now, and wouldn't >>> pay a lot for a dedicated box. >> >> He asked an on-topic question and got shitty, off-topic answers, >> all of them but maybe 2 or 3. >> Your's was none of the 3. >> >> Gerhard > > I answered first, and my response was civil and relevant. He didn't > answer my question about schematics. > > He apparently resents not getting free consulting, and resents > discussion drift. It's an unmoderated public forum; get used to it. > >He doesn't even understand how Usenet works but has "expectations"
Reply by ●October 30, 20182018-10-30
On 10/29/2018 09:25 PM, jjhudak4@gmail.com wrote:> I did reply to your first post but I forgot and did it in my email and it didn't get posted. To address your points: > 1) I usually try to find schematics, sometimes not very successful. If I do have schematics, I try to model it to see how it performs (if it is something I am not familiar with). I did answer the question of noise. > 2) There are lots of reasons why curve tracers are no longer made by Tek, HP, etc. My speculation is that as a business decision, it is not profitable because ppl solve the design problem with different technology rather than with discrete transistors. That does not imply the use/application is useless. > 3) I know a curve tracer does not provide any insight as to noise characteristic, and I know that noise is a concern - especially in old clunky electromechanical beasts like cassette deck. If you look at some of the designs and implementations, some of the low noise solutions are quite elegant. > 4) Discussion drift - that is true, I resent it. I ask a technical question on a technical group and expect the answers to be technical. As one reader pointed out, most of it was not. > 5) free consulting - is appreciated, and thanked in kind. Just because an approach doesn't do something doesn't mean the approach is invalid for other things. Because curve tracers do not provide insight on noise characteristics doesn't me they aren't useful for their intended purpose. I have a fairly solid academic background in EE but admittedly rusty in certain areas. I took extra courses in semiconductor physics even though it was not my major and that was many years ago.You "resent" it? You "expect" answers to be what now? You show up like a noob from Google Groups five minutes ago with a bunch of demands dropping your credentials like anyone care. Who is even you?> 6) unmoderated public forum. Right - anybody can say anything on any topic. My expectations for ppl to stay on topic was apparently incorrect. My expectation of professionals providing feed back in a professional manner was not met by a number of the responses, to which I am unpleasantly surprised. I am used to a higher standard, even in unmoderated groups. My expectations have been 'recalibrated.' > I appreciate the helpful feedback from you and other. > Peace, > JOMG get a load of this guy. Get over yourself, sweetheart. You don't "own" a thread you start and nobody was even talking to you when it went off-topic.
Reply by ●October 30, 20182018-10-30
three_jeeps wrote:> > > Give it a rest man. I have some high end audio gear from the 80s and > I do some repair for friends/clients. >** Would that include amplifiers made by Phase Linear, SAE, Crown, BGW, Yamaha etc ? Worked on quite a few of them and can say what many of the house or "secret" part numbers are. .... Phil .... Phil> I don't care about your technology views, your likes/dislikes, or your perspective on audio or electronics, your political view, or what your father did or did not do. Mine was a WWII vet - fighter pilot but has nothing to do with the question I asked. > > Why a cassette deck? I am repairing one for a friend. He and I both have a Nakamichi 1000ZXL. It appears his preamp died. I didn't challenge him as to why he *still* uses it...to each their own. > take a back seat, shut up, and enjoy the ride, or leave.
Reply by ●October 30, 20182018-10-30
On Sunday, October 28, 2018 at 6:26:53 PM UTC-7, bitrex wrote:> On 10/28/2018 08:46 PM, jjhudak4@gmail.com wrote: > > I am repairing some high end audio amps and cassette decks that have house numbered semiconductors, mainly transistors as well as JDEC p/n that are NLA.> > Basically what I am looking for is a curve tracer that has ...> My possibly-biased-by-youth opinion is that the transistor curve tracer > is a mostly obsolete instrument from a time when transistors were > expensive and manufacturing tolerances sucked balls and > duff/under-performing parts were commonplace, that's not year of our > Lord 2018, discrete transistors are cheap and quality is excellent.Though the function of many devices is, as you say, excellent, yet a few hours with a curve tracer is an important piece of experience to have. Watch a semiconductor overheat, make smoke, and then... see how it died (or how it recovers and continues to function when cooled). When a transistor fails, it often passes the two-diodes testing, but a curve trace that shows it to have 6V breakdown instead of 80V, is giving you priceless information.> Like maybe if you needed to well-characterize RF or microwave amp > transistors to get all the relevant information...Yes, a curve tracer is an exploratory tool, and gives good service at the borders of the known. That may be RF, or it may be Zener noise, or it may be Early effect, or any of the characteristics that a design (or repair) will hinge on. Specification sheets just do NOT cover everything, it takes some exploration too. Don't leave the well-blazed trail without your axe, compass, and curve tracer.
