Add GW Instek to the good guys list

For vendors who sell new stuff.
Forum rules
This is for links to mainstream, long established vendors you like -- or you don't, please comment on the service you've gotten from them in the past either way. Ebay stuff goes on surplus subforum, even if they are an "official ebay store" as most of those are more or less surplus vendors who cannot get more after they run out of something -- even if what they are selling is "new".

Add GW Instek to the good guys list

Postby Doug Coulter » Wed Jan 30, 2013 1:39 pm

OK, this was all "my bad" but I managed to let a full output arc from my Spellman (also on our official super good guys list) into a scope probe due to an arc inside the tank to something the probe was on. This is my super-nice GDS 2204 DSO scope. The probe is toast, but before it opened up, looks like it arced through and fried the CH2 input board in the scope.

Fried the probe. Fried the scope channel - 4 resistors no tops. Most parts house marked. I hadn't registered for the "limited warranty" and heck, it was my bad anyway.
I tried calling the dealer I bought it from. They can't get spare parts. They sent me to the phone number for GW-America, which works, and I got hold of a real tech.
After trying to understand one another for awhile (he was obviously oriental), he just gave up and sent me the schematic for that board!

Really, that's almost unheard of. Only Spellman has ever done that for me in the past, after I agreed to not share their trade secrets (a promise I've kept). Now I don't have to guess what all the house-numbered tiny smd parts are, and can incircuit check things like protection diodes, and order new semi's as required - the schiz has real part numbers on it.

This makes me very happy! I can now fix this expeditiously right here, no shipping the thing back and forth and all that agony, along with labor and parts charges. All the fried parts appear to be simple stuff I have in stock or that is dirt cheap - resistors.

Note, I'm going to return the favor. Their email address listed for GW-Instek America does not work - it bounces with a 505, no relay allowed (presumably because it relays to Taiwan or something). But now I have the tech's email address, so I can tell them about it.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Add GW Instek to the good guys list

Postby johnf » Thu Jan 31, 2013 2:10 am

Doug
Most excellent
not often that you can get through the pit bulls at the door, and even better they have come across.
It does happen but sadly not often these days as I'm sure that many have been burnt by their product coming back copied by those who can not design theirway out of a wet paper bag.

Go GW Instek
johnf
 
Posts: 433
Joined: Sun Aug 08, 2010 3:51 pm
Location: Wellington New Zealand

Re: Add GW Instek to the good guys list

Postby chrismb » Thu Jan 31, 2013 10:10 am

Doug, you wally! Make sure you've got a TVS across the probe if you have capacitively coupled HV flying around! Maybe you can build it into the probe somehow.

...Maybe there's a market for BNC adapters that contain a TVS - never thought to look to see if someone sells such a thing already?

I routinely put TVS across the outputs of my power supplies now, following the PKR failure that I reported repairing.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Add GW Instek to the good guys list

Postby Doug Coulter » Thu Jan 31, 2013 11:12 am

Thing is, a transient suppressor has some serious parallel capacity - not what you hang across a 1 meg/200 mhz scope input if you want to keep the frequency response.
The solid state ones add some serious DC leakage as well. So, you're kind of messed up there for scopes - it would work for lots of other things, though.

In this case, I had a direct arc of main HV to a faraday probe. That probe had a kludged up 10 meg 100::1 divider after it, before the scope, in discrete 1M/100k half watt resistors. That arced over too, then the probe, then the junk inside the scope. A regular TVS would just have added to the smoke in this case - that's a lot of joules.

It looks like a bunch of little series R's - 0603 smds, in the 100 ohm range, protected the protection diodes in the thing. I might have one blown FET or opamp too - dunno yet, but the R's are on the way from digikey - do the obvious things first! At least no PCB tracks were fried. I guess the pulse wasn't very long.

I have figured out how to programmatically capture scope data from the GW scopes I have (same protocol as tektronix) in more or less real time. What this means in a practical sense is that at least for short bursts I can effectively get data sampled at 1 ghz into the computers for later analysis - and in time-sync across all the scope channels. These scopes capture a lot more data in a burst than they show on-screen, like about 1-2k samples/ch, and tell you which one the trigger occurred at.

Some interesting implications there - I could make a MCA out of a scope here - even all 4 channels independently should I want to. Wouldn't be as good as the purpose built ones, but...price is right, and it's probably "good enough". A heck of a lot nicer than trying to low pass input pulses to the point of digitizing them with a soundcard at 48 khz, which can't handle fast count rates (pulse pileups in the lowpass filter). I'll be looking at the software for that soon - hopefully it won't be real hard to get going.

The real bummer in events like that is I lost the rest of the data from that run - this is why I'm writing software for internal lan name resolution. In the future, the database will be on another machine, via wireless, so safe up to the point of an incident, even if the collection machine down by the fusor is fried or crashed. So, I'll have info on the conditions just before the "issue". As it is now, using just log files down there - if the file isn't closed and the computer crashes - it's usually lost. So I'm upgrading to a "real" database on a remote machine, slowly but surely.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Add GW Instek to the good guys list

Postby chrismb » Thu Jan 31, 2013 11:49 am

Doug Coulter wrote:Thing is, a transient suppressor has some serious parallel capacity - not what you hang across a 1 meg/200 mhz scope input if you want to keep the frequency response.


OK... I can't really argue with that, if you are wanting to see to 200 MHz.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Add GW Instek to the good guys list

Postby Doug Coulter » Thu Jan 31, 2013 2:39 pm

I like your idea in principle - just not for this one app. For a whole lotta things, you can find decent transient-getters, and your idea to make an inline BNC one is not a dumb idea. I have several cases around here where some little bit of extra load won't hurt - in other data aq.

However, we skipped those in the data aq design we're coming out with soon (you guys are gonna love it, or so I hope). This is because the PIC chip inputs are 100's of megs, so series R doesn't affect a/d accuracy much at all - yet they will take a few hundred ma if you try to force the inputs past the rails. At that point, a very simple high valude RC network means that even kilovolt levels can't fry anything but a good, old fashioned, 10k through hole resistor - which is obvious and easy to replace if you have an oops.
Use a low voltage C, say a 3uf 5v MLCC after that resistor, and it IS a TVS in effect.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA


Return to Mainstream, new stuff

Who is online

Users browsing this forum: No registered users and 3 guests

cron