What's the general consensus on Xicon Carbon Film resistors?

[nevetslab]

I just got a note from the Tech at Hiwatt, that they are changing from the TE Connectiviry CFR100 series 1W Carbon Film resistors (look just like the Mepco/Philips parts that they've been using for years). to Xicon parts. I've always had Xicon at the bottom of my list to use only as a last resort. But maybe it's just my attitude about that mfgr's parts. I'll order some to have a look.

[Chuck H]

I haven't used them in a few years but never shirked them when ordering for a build. I do know I've used them though I can't recall with certainly which projects. Four years ago? Ten? That said, I never had any problems with them at all. I recall testing their value for particular circuits and finding their accuracy specs to always be far better and the published allowance.

I never did this, but if you're really concerned you should compare ALL data sheet specs (voltage rating, temp rating, temperature value derate, etc.) to the TE CFR resistors. But as far as consistency and accuracy I don't think you'll have any problems.

[Enzo]

Did the tech note state WHY? In the world or OEM, it isn't always about the part itself. It could be as simple as the one brand can reliably guarantee the million-part orders whereas the other brand cannot.

[nevetslab]

I believe the client is changing due to parts not reliably being available for production qty's. Today, after getting my Quan Tech 2500 Automatic Resistor Noise Test Set up and running, and went thru the small qty's of TE-Connectivity CFR100 series 1W Carbon Film resistors, I ordered parts from Mouser in the Xicon CF-RC series 1W Carbon Film resistors to compare them with the CFR100 series parts. The test set runs them at specific DC current levels based on the resistor wattage/value where DC Voltage applied is = sqrt Res Value x 0.25W for 1/2W & larger, or x 0.1W for 1/4W and smaller. Voltage applied up to 250VDC. Short term applied current, then noise voltage thru 1kHz BW filter, amplifier, RMS rectified, S/H and displayed on a 40dB range meter with user-setable display time. One of the interesting lab instruments I bought decades ago during product development at BGW Systems.





This instrument came from a MIL-STD-208B, Method 308, which was written around an earlier Quan Tech 315 Test Set. I read thru it, having found Bill Perkins of Pearl Hi-Fi in Canada had parked a pdf of the standard on the internet. Then found a copy of the earlier Quan Tech 315 manual to see if I could figure out just how to predict what a given resistor's Noise Index would be over this -40dB thru 0dB dB Meter. Still don't know. I've always used it empirically to get the nominal reading range, then set the two threshold thumbwheel switches to yield a 5eB window and then go thru the batch of resistors looking for consistent readings and weed out those that are noisy. No idea what I'll find with the Xicon parts. Interesting that their data sheet showed a noise current graph, while TE Connectivity's CFR series resistors didn't even mention noise attributes of their resistors.

I did at least compare the CFR series with same-value Metal Film resistors and found consistent readings with both. That of some Carbon Comp resistors read higher noise.

Current_Noise_Test_for_Fixed_Resistors.pdf

[The Dude]

I see the resistor under test is in a metal box- presumably grounded/shielded. I'd still be curios to know what effect surrounding noise has on readings since we're talking about such low level noise. Have you ever tried holding a cell phone or other possible noise sources nearby while testing?

[Helmholtz]

Have you really seen resistor noise problems in a tube amp with (non-defective) CF types?

The main noise sources in a tube amp typically are the input triode's grid stopper and the input triode itself.
As no current flows through the grid stopper, it's only thermal noise which depends on value and temperature but not technology/construction/type.

[nevetslab]

Usually it's Plate Resistors that get noisy. This episode of checking these components is mostly my biased distaste of Xicon resistors....though I can't really say why that is? With the client making a blanket decision to depart from what I've always considered a top quality (TE Connectivity, which look like the latest owner of what was previously Mepco and then Philips) and changing over to Xicon, I thought I'd see if anything shows up with this particular evaluation. I can't say for certain that over the years I've had the instrument, that it really proved useful. Just thought I'd have a look. Last time I used it was to evaluate different mfgr's 1/8W resistors being used in a Hydrophone Preamp board I was building for towed arrays.

[Helmholtz]

Usually it's Plate Resistors that get noisy.

Yeah, but doesn't that look more like a reliability issue requiring some time and stress to develop?
Is it predominantly with Xicon resistors?

In my company we considered Beyschlag (No.1), Roederstein and Philips resistors top quality for low power.
But that was years ago. Didn't follow up what became of those brands.

[nevetslab]

I don't believe I ever came across the Beyschlag resistors here in the US. Roederstein...yes, but only with their film caps. Don't recall seeing their resistors. The Philips brand....I think that was Mepco before, then I believe Philips bought Mepco, seeing both names on their thick component catalog. I may still have one of those in my lockers. At BGW Systems, we used their 1/2W carbon film resistors almost exclusively, other then when they were out of stock with our suppliers. I think TE Connectivity is who absorbed it at present. Not sure who it was before that.

This afternoon, I moved one of our Aguilar DB751 Bass Amps up to the bench, which had excessive tube noise issues. It has three tubes...front end, buffer/Effects Send stage and final stage. 420VDC supplies, then dropped slightly with 10k/filter and 100k plate resistors, 1.5k to 1k cathode resistors depending on where. After getting nowhere with tube substitutions, I looked at what looked like Allen-Bradley or Ohmmite Carbon Comp 100k 1/2W resistors, and decided to remove the PCB from the chassis and replace those with new ones I had in my parts drawer. I still had the Quan Tech set up for 100k value, and used it to select four new resistors. Swapped them out, and prior to putting the PCB back into place, I decided to tack-solder lead wires onto the freshly-removed resistors from the amp. Noise level on those were a good 10dB or more higher than what I just installed.

Powered the amp back up, and it was back to normal....enough so that I was able to put the original EH 12AX7 tubes back in....one stage at a time. It's still burning in, in the background to be sure all remains ok.

There's a switch on the instrument that lets you read the System noise with the selected resistor in place in the shielded fixture, then you can read total noise, which adds the current noise of the resistor with the DCV applied. With those parts I removed, the difference between the self-noise using those resistors and then applying the DCV to them, one was 30dB greater, two were around 18dB greater, while those I replaced them with, the difference was only around 7db.

Never used this box before in this way, but it does seem to have aided in finding the noise problem on the Aguilar amp. I did at least get to restore the tubes that I thought were the problem.

[Helmholtz]

Though there are different qualities, carbon comp resistor are notorious for current distribution noise.
I assume that the noise level increases with age as the value drifts upward.
But only plate resistors close to the input (with a lot of gain to follow) should really matter.

[nevetslab]

I found this when I googled Mepco Electronic Components https://www.procureinc.com/manufactu...ents_-_Vishay/

Mepco did get acquired by Philips in 1999, then they spun off the passive components business at BC Components (Beyschlag Centralab Components). Vishay then bought them. No referrence to TE Connectivity, so maybe it was just a guess, as the CFR series resistors looked just like the Mepco/Philips parts I was used to.

[tedmich]

I can confirm the plate resistors of several high gain amps are spec'd as metal film even by cost conscious companies like Peavey. Their JSX notably has all MF plate resistors in the preamp and PI, while every other resistor in the amp is CF (except (2) 5.62k resistors in the bias circuit, maybe unavailable in CF)

They do seem to have made some good component choices for this model; with typically cheap ceramics in the tone stack switched to NP0/C0G dielectric. And the MF in the preamp/PI cost a whopping $0.005 - $0.004 when the pdf was made, as clicking on the component shows.

And for my applications Xicon while quite cheap is often perfectly acceptable!

[Leo_Gnardo]

There's a switch on the instrument that lets you read the System noise with the selected resistor in place in the shielded fixture, then you can read total noise, which adds the current noise of the resistor with the DCV applied. With those parts I removed, the difference between the self-noise using those resistors and then applying the DCV to them, one was 30dB greater, two were around 18dB greater, while those I replaced them with, the difference was only around 7db.

Very instructive! You got the gadgets nevets, and it's good to see this one put to use.

[nevetslab]

I can confirm the plate resistors of several high gain amps are spec'd as metal film even by cost conscious companies like Peavey. Their JSX notably has all MF plate resistors in the preamp and PI, while every other resistor in the amp is CF (except (2) 5.62k resistors in the bias circuit, maybe unavailable in CF)

They do seem to have made some good component choices for this model; with typically cheap ceramics in the tone stack switched to NP0/C0G dielectric. And the MF in the preamp/PI cost a whopping $0.005 - $0.004 when the pdf was made, as clicking on the component shows.

And for my applications Xicon while quite cheap is often perfectly acceptable!

I think the Mesa amps....Lone Star, Dual and Triple Rectifier Solo use Metal Film resistors there. Might be RN70's which have the higher wattage and Working Voltage rating. I'd have to look and see if I have photos to verify that. I did find images inside the Mark V chassis:



They're RN70-size, from what they look like. All Metal Film 1%. They always did use great quality components in their build, though still a PITA to work on.

Earlier today, I was restocking my 1/2W Carbon Film drawers over on Digi-Key's website. Stackpole CF12 series 1/2W mostly. Price was right...25 pcs for $1.25 per value. I had added a 100k 1/2W Ohmite part, then when I went to pay for it all, there was an additional $10 shpg fee from Ohmite, and I deleted it, thinking I'd find another source. That was a pipe dream. Most other sources, including Just Radios was like $3.95 per 5 pcs...with a $30 min order plus the shpg from Canada. Antique was about the same price, The biggest joke was trying to find them on ebay. I went back to Digi-Key and paid $12.76 for 25 pcs.

It might have been the Xicon 5W Square Cement filled power resistors that seemed to fail too easily I recall the days in the BGW Systems lab, putting various vendor's 5W bathtube emitter resistors in the vise and breaking them apart to look at the construction. Quite a wide variation of what's hidden side those parts.

[nevetslab]

Very instructive! You got the gadgets nevets, and it's good to see this one put to use.

A couple days ago, when I first had this up and running, remembering some years back, I had ordered power supply caps for recapping the HV PCB assy and the LV PCB assy. Looked for what I had bought, but lost track of them, and powered it up. I did have to open it up to replace the power switch, and that was an exercise in just HOW DO YOU gain access to the innards. A couple wrong guesses before success, then replaced the SPDT toggle switch. Put it back into the housing, and read thru the manual to recall how to run the beast. A short while later, smoke came billowing out the top front gap in the panel, as I switched it off. Taking it apart, I found the HV PCB's two 30uF/450V Axial Sprague Atom Electrolytic caps had burst, spilling electrolyte all over the PCB and down into the card guide and onto the floor below the card cage. Got all that cleaned up, removed the two damaged caps, and also pulled the LV PCB assy out as well, it having four 500uF/50V Axial Sprague Atom caps as well, though they weren't warm. Those were wired in series, without ballast resistors. I did check those out on my bridge, and DF was a little on the high side for having date codes of 1978. I ordered replacements thru Just Radios for all that. I did at least have a pair of 47uF/450V Axial caps for the HV supply card.

That instrument WAS helpful while working on that Aguilar DB751. I hadn't made use of that System Noise/Total Noise switch in the past, but that IS mentioned in the Mil Standard 308, comparing the System noise (which has the unpowered selected resistor being measured by the measuring amp/filter system) and then comparing that to the Current Noise found when the specified DCV is applied. Seeing the VAST difference on those four resistors I removed from sitting underneath the 12AX7 tubes in the Aguilar...nearly touching the glass....and that of fresh Carbon Comp resistors I had.....THAT was an eye opener! It was after putting the PCB back into place that reality struck me. I placed the new resistors back onto the component side, in the same close contact position with the bloody tubes! I should have placed them on the back side of the PCB, and elevated them further from that heat source. I have another dB751 coming over for service, so I'll have another opportunity to do that. I also sent a note to Dale at Augilar's Customer Service regarding those plate resistors getting cooked by their tubes sitting horizontal like that. Clever packaging....but not well thought out.