I'm sure we've all repaired some equipment where a tantalum capacitor has shorted.?
?
I lived for 30 years in San Jose, and collected lots of parts, including a lifelong supply of teardrop tantalum capacitors of various values and voltages. I have used these in projects over the years with no trouble. But over the past few years they seem to fail immediately. Today I had to replace a tantalum capacitor in an HP 5328A counter, and the replacement immediately failed. I placed another one on a power supply and ran the voltage up to 15 volts for a 20 volt capacitor. The supply briefly went in to current? limiting multiple times before the capacitor failed completely.
?
Do tantalum capacitors degrade with time with no voltage supplied? Or do I just have a bunch of poor quality parts?
?
Any information would be appreciated - although regardless, I think it may be time to toss my collection of tantalum capacitors.
?
Gary Appel
|
Re: VNA Resonance Methodology
On Wed, Nov 27, 2024 at 11:05 PM, Tom Lee wrote:
Thanks very much, jmr, for going to the trouble of running the numbers through Genesys. You've provided a valuable cross-check of my cheese formulas, which give L = 64uH, Q = 190, and srf = 16MHz. The inductance comes out quite close (thanks to Wheeler). The Q formula is extremely crude (assumes resistance for close-wound coils is 4x the textbook skin resistance), but seems to be in the right general ballpark, if optimistic. The srf formula comes from simplifying a helical transmission line formula. From your curves, I infer that it's giving a reasonable approximation.
Thanks. It's a long time since I wrote that software program but (to compute Q) I think the program uses a map based on empirical measurements made decades ago by Butterworth and Medhurst some time around 1947? Both noted that the AC resistance changes from the skin effect model of a straight wire and they produced tables of how it changes based on the L/D ratio of the coil and also on the wire to gap ratio. So I plotted all this data in a huge 3D map in my software and the program code then interpolates between the data points as required. I can't remember whose lookup tables I eventually used, it might even have been a later revision of the data I found in a modern text book. I do the same for the impact of the self capacitance of the coil in that I mapped all the data for this as well.
?
Before this I had to do it all in an excel spreadsheet and manually navigate the tables to find the various fudge factors for Q and capacitance etc.
?
The Genesys air solenoid model seems to work quite well although for wideband stuff at work I prefer to use a VNA based model or a physical model I developed at work many years ago. The Genesys model (and my software program) falls apart near the first resonance mode so it is hopeless for wideband design work. My physical model is transmission line based although it is a bit more involved than just a transmission line. I did all this at work many years ago in order to be able to understand how to make better (ultra wideband) bias tees and how to make RF filters work over wider bandwidths (especially highpass filters). The aim was to be able to model the various resonance modes in a solenoid such that the model could be used in a simulator.
?
?
|
Re: VNA Resonance Methodology
Thanks. It is a bit hard to see the blue arc in the image so I've thickened up the blue trace width and plotted it again without the markers. The top of the blue arc corresponds with the start frequency of 1 MHz and the bottom of the arc corresponds with the stop frequency of 2 MHz.?
?
1.45 MHz is roughly at the centre of the arc at 3 o'clock on the smith chart (very close to being an open circuit).
?
|
Re: HP 6114A - looking for R8A, 15.82K 0.1% 1ppm resistor
Maybe because I also wrote them?...?
The ones I ordered before - bulk foil Z201 - are not smd, though, and I wonder if that has to do with them coming up with an MOQ (the VSMP1206, which is what I think you ordered, are quite a bit cheaper). I inaccurately assumed you're looking for a through hole type.? But hopefully they reassess their stance and do this for you. Radu.?
I asked for the 1206 size as I only needed 1/8W and the price was less.
Here's the part ?
?
Looks like I did get a followup email after declining moving forward saying "Let me ask manufacturing as maybe the unit price will be higher just for one."
?
Dave
?
|
Re: VNA Resonance Methodology
Thanks very much, jmr, for going to the trouble of running the numbers through Genesys. You've provided a valuable cross-check of my cheese formulas, which give L = 64uH, Q = 190, and srf = 16MHz. The inductance comes out quite close (thanks to Wheeler). The Q formula is extremely crude (assumes resistance for close-wound coils is 4x the textbook skin resistance), but seems to be in the right general ballpark, if optimistic. The srf formula comes from simplifying a helical transmission line formula. From your curves, I infer that it's giving a reasonable approximation.
Thanks again!
-- Cheers
Tom
--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
420 Via Palou Mall
Stanford University
Stanford, CA 94305-4070
toggle quoted message
Show quoted text
On 11/27/2024 1:59 PM, jmr via groups.io wrote: If it helps add some confidence to this stuff, I used the air cored solenoid model in Genesys to predict the inductance and Q vs frequency for your coil. 49 turns 31mm diameter length 22mm wire diameter 0.4mm.
This model in Genesys is typically quite good and I'm lucky to have access to CAD tools like this. Genesys predicts 65uH and a Q of 122 at 1.45 MHz as in the screen capture below.
|
Re: VNA Resonance Methodology
Richard,
It is a short blue arc at the right edge of the chart.? Pretty hard to see unless you save the image and enlarge it.
--John Gord
?
On Wed, Nov 27, 2024 at 02:48 PM, Richard Knoppow wrote:
I see zilch on the chart. Did something get left out?
On 11/27/2024 2:16 PM, jmr via groups.io wrote:
Here's what you should see on a smith chart with a start frequency of 1
MHz and a stop frequency of 2 MHz when resonated at 1.45 MHz with a
187pF cap and with a series 50R resistor inline.
--
Richard Knoppow
Los Angeles
WB6KBL
SKCC 19998
|
Re: VNA Resonance Methodology
I see zilch on the chart. Did something get left out?
toggle quoted message
Show quoted text
On 11/27/2024 2:16 PM, jmr via groups.io wrote: Here's what you should see on a smith chart with a start frequency of 1 MHz and a stop frequency of 2 MHz when resonated at 1.45 MHz with a 187pF cap and with a series 50R resistor inline.
--
Richard Knoppow
Los Angeles
WB6KBL
SKCC 19998
|
Re: VNA Resonance Methodology
Here's what you should see on a smith chart with a start frequency of 1 MHz and a stop frequency of 2 MHz when resonated at 1.45 MHz with a 187pF cap and with a series 50R resistor inline.
?
|
Re: VNA Resonance Methodology
If it helps add some confidence to this stuff, I used the air cored solenoid model in Genesys to predict the inductance and Q vs frequency for your coil. 49 turns 31mm diameter length 22mm wire diameter 0.4mm.
?
This model in Genesys is typically quite good and I'm lucky to have access to CAD tools like this. Genesys predicts 65uH and a Q of 122 at 1.45 MHz as in the screen capture below.
?
Although the red Q curve looks very similar to the Q curve in my earlier software plot (with a peak of just over 200 at about 7.5 MHz), I wouldn't trust either Q curve beyond about 80% of the way up the rising part of the curve. Above this I think both my software and the Genesys model will have creeping errors that could become very significant. Fortunately, you are operating down at just 1.45 MHz and I think the Q estimate of 122-124 across both tools should be fairly close to being correct.
?
When making the coil for real, there will be uncertainties in the winding accuracy and the wire diameter and the enamel thickness of 'real' 0.4mm wire and these will introduce some uncertainty. Plus there will be some inaccuracy in the software tools (but I think not much down at 1.45 MHz in this case).
?
?
?
 ?
|
Re: HP 6114A - looking for R8A, 15.82K 0.1% 1ppm resistor
I asked for the 1206 size as I only needed 1/8W and the price was less.
Here's the part ?
?
Looks like I did get a followup email after declining moving forward saying "Let me ask manufacturing as maybe the unit price will be higher just for one."
?
Dave
?
|
Re: VNA Resonance Methodology
If I go with what you posted earlier I can assume that your current tank coil has 49 turns on a 31mm OD former using 0.4mm diameter wire (close wound) I can punch these numbers into the software program.?
For 0.4mm enamelled wire I usually assume about 0.06 to 0.08mm is added to the diameter of the wire because of the enamel coating. The enamel coating means the copper itself isn't perfectly close wound.
?
So I've set the centre to centre gap of each winding at about 0.46mm in the software program.
?
The software then predicts an inductance of about 66uH at 1.45 MHz and a Q of about 125. See the image below.
?
This software is based on about 75 year old equations and lookup tables for inductance and Q factor etc. These old equations and tables have been proven many times by lots of people to be reasonably accurate.
?
So I would expect your inductor to have an inductance of about 66uH and a Q of about 125 at 1.45 MHz. It shouldn't be far off these numbers unless the former you are using is adversely affecting the coil.
?
A Q of 125 isn't that great. It should be possible to achieve a much higher Q if you use thicker wire and change the coil dimensions. This will end up having much larger dimensions than your current coil though. I can show you what I mean using the software if that helps?
?
The alternative is to use Litz wire. Something like 160 strands of 48 gauge wire for example. This would give a much higher Q although it won't be cheap to buy Litz wire like this. It might double the Q compared to using a single strand of copper wire of equivalent diameter.
?
?
?
?
|
Re: HP 6114A - looking for R8A, 15.82K 0.1% 1ppm resistor
That's surprising to me - I hope they're not changing their business model. I've been able to order single parts at my exactly specified value to two decimals before, on the Z201 variety. Is this what you asked for??
I suggest persisting. I can pull my records and see who you talked to vs. my contact (let's work on this offline, Dave). I stated to them I appreciate them doing this, which they responded to, so there's a conversation to reference.? Radu.?
I contacted TTC and the reality is less exciting than what their website suggests.? Their reply was "The MOQ for the series is 5 pc. since is a very peculiar R value."
Not sure why my 'any custom value from 250 to 25K' is special.
?
Meanwhile I can find 5ppm parts of 15K and 820 ohm resistors in stock from Digikey for <$5 shipped, and really I just need to be in the range that A2R5 can adjust for so I could just use a single close value vs hitting 15.82K exactly.
|
Re: HP 6114A - looking for R8A, 15.82K 0.1% 1ppm resistor
I contacted TTC and the reality is less exciting than what their website suggests.? Their reply was "The MOQ for the series is 5 pc. since is a very peculiar R value."
Not sure why my 'any custom value from 250 to 25K' is special.
?
Meanwhile I can find 5ppm parts of 15K and 820 ohm resistors in stock from Digikey for <$5 shipped, and really I just need to be in the range that A2R5 can adjust for so I could just use a single close value vs hitting 15.82K exactly.
|
Re: VNA Resonance Methodology
Hi jmr,
Well, it's an interesting methodology for sure. I do have a Nanovna which I bought purely for evaluation purposes, but I have two proper lab-grade HP VNAs as well. Can't recall the model numbers off hand but one's good to 1.3Ghz and the other to 6Ghz (expanded from 3). Unfortunately the matching S parameter test set? for the 6Ghz one tops-out at 3Ghz, however, which is a bit of a nuisance. Still, since we are only talking 1.4Mhz (virtually DC :-)) for this test, none of that fancy stuff is necessary anyway. :-)
Jinxie
|
Re: VNA Resonance Methodology
Thanks, Froggie. By all means PM me when you have access to all your test equipment again.
|
Re: Sticky ink of schematics
Personally, I have found that the laser printer at home doesn't
do this while the ones at work and in the print shop I use do ---
the difference may be that the one at home is black+white only and
is of the black powder toner+fuzer variety while the others are
colour and use a waxy sort of ink.
Luckily, when I needed some schematics printed on A3 I found a
print shop that had a black+white plotter that could go up to A0
and those haven't stuck to anything (yet).
I did at one point spray some colour prints with a clear lacquer
which seemed to work but a PITA.
As Ed says, Inkjet is the best alternative.
Dave
On 27/11/2024 06:26, ed breya wrote:
toggle quoted message
Show quoted text
Like Chuck said, laser printed items tend to stick to other
stuff over time. I haven't seen it too bad from one paper sheet
to another, but definitely from paper to plastics, or (toner
sides) face to face. I've seen and heard it (sounds like peeling
Velcro apart) especially from laser printed paper pages to
(inside) vinyl covers of 3-ring binders, and plastic page
separators and such, leaving some of the print transferred.
?
I think the best way to preserve paper documents is to laser
print one-sided only, and avoid any long term contact with
plastic materials - an extra blank paper cover sheet on top can
isolate it from a plastic layer. If it must be two-sided, go
with inkjet instead.
?
Ed
|
Re: HP 8341B 1992 firmware version
That is correct, the firmware is the same for both. I have used exactly this firmware to update my 8341B without issues.
|
Re: HP 8594E PSU repair/debug safety checklist
Hi Group check the power switch they tend to fail on those.
if the PSU is faulty contact me on OZ11LPR@... I have alot of parts for those units.
It is a pain to work on that powersupply type.
best regards Peter
Differential probe isn't going to keep *you* safe......? The isolation
transformer might.? (And most variac's are NOT isolation transformers)
On Tue, Nov 26, 2024 at 9:53?AM Ke-Fong Lin via groups.io
<anotherlin@...> wrote:
>
> Hi everyone,
>
> My HP8594E has a temperamental PSU, sometimes everything just works fine otherwise the SA just doesn't turn on.
> Something is obviously wrong and needs to be debugged.
>
> I've not opened the case yet. I've just searched a bit the internet ().
> HP doesn't provide schematics but people have already attempted repair.
>
> I may have to work on the "mains connected" (120VAC US or 220VAC Europe) side.
> So I just want to be sure that I would be safe with a differential probe, and if I use a diff probe, an isolation transformer wouldn't be needed.
>
> Seems that "curious marc" is using a variac as some form of isolation transformer:
>
>
> In general, do you guys have some links or checklists for safety when working on these kind of switch mode power supply.
> Thanks!
>
>
> Best regards,
>
|
Re: Sticky ink of schematics
Like Chuck said, laser printed items tend to stick to other stuff over time. I haven't seen it too bad from one paper sheet to another, but definitely from paper to plastics, or (toner sides) face to face. I've seen and heard it (sounds like peeling Velcro apart) especially from laser printed paper pages to (inside) vinyl covers of 3-ring binders, and plastic page separators and such, leaving some of the print transferred.
?
I think the best way to preserve paper documents is to laser print one-sided only, and avoid any long term contact with plastic materials - an extra blank paper cover sheet on top can isolate it from a plastic layer. If it must be two-sided, go with inkjet instead.
?
Ed
|
Re: HP 8341B 1992 firmware version
There is no differences in CPU board and firmware between 8340B and 8341B.
Calibration constants 79 and 98 must 2 for 8340B and 5 for 8341B, according to service manual.
|
Jinxie