Resistor types: identification.

[mhennessy]

Yes, it's well-known that film resistors have a voltage rating that is in addition to that implied by their power rating.

But it's perhaps less well-known that even if you stay within these ratings, higher-value resistors that are subject to high voltages will eventually drift high and ultimately fail O/C - I use the "rule of hundreds" as a guide (>100k, >100V), but I've known it happen at 47k/50V.

I've explained this elsewhere (doubt I'd be able to find the post, if it was even on this forum?), but suffice to say, it's a very common cause of failure of start-up resistors in switched-mode power supplies, amongst other things.

I've never seen any comment from manufacturers about it; I'm not saying none has ever written about the subject, but if they have, I've never stumbled across it.

For what it's worth, metal oxide resistors are much more dependable at high voltages.

[Lucien Nunes]

IIRC Microvitec used a carbon comp startup resistor in their popular 'Cub' monitors, every other resistor in the chassis being film. But they went open circuit with monotonous regularity anyway.

[SteveCG]

Re post no.18:

Al, now you're asking !

I think it was centred on 45 Mc/s for quite a few of the radio-telescopes (originally based on the war-time Pye radar IF strip - which in turn was based on the pre-war Ally Pally TV transmissions). However the later 5 Km telescope may have had a higher IF frequency. All used carbon composition resistors as an explicit design decision.

[Skywave]

Originally Posted by mhennessy

For what it's worth, metal oxide resistors are much more dependable at high voltages.

Which is what I have very recently read elsewhere: metal oxide film resistors are more reliable at high voltages than 'ordinary' metal film types. For maintenance work on kit that uses HV, an ability to discriminate between each type is now clearly very useful. And that brings me back to where I came in: how to identify - visually, or otherwise - the difference between those two?

Al.

[Skywave]

SteveCG: your post # 23. My earlier post - in which I mentioned that I did not detect any unwanted inductive effects in the attenuator I had repaired - was with that attenuator connected to a resistive termination, of course. Thus the effects of any stray L would have been relatively low. However, in a 45 MHz IF strip with its tuned L/C ccts., I can well believe that the inherent inductance of spirally-wound resistors could indeed render them quite unsuitable.

Al.

[mhennessy]

Originally Posted by Skywave

metal oxide film resistors are more reliable at high voltages than 'ordinary' metal film types. For maintenance work on kit that uses HV, an ability to discriminate between each type is now clearly very useful. And that brings me back to where I came in: how to identify - visually, or otherwise - the difference between those two?

It's just a question of being familiar with the families that exist. In the old days, the pictures in the RS and Farnell catalogues were a great source of data - and I wish I'd kept a few of the older ones! Today we have the internet, but although there is much more data out there, browsing and searching isn't perhaps as easy as it once was.

That said, a quick "Google Images" search for "metal oxide resistors" brings up nearly all the types that I'm familiar with. Same for the other resistor types.

Regarding your original question (identifying between metal film vs carbon film in an item that you are repairing), I don't really understand why there is cause to worry. Simply replace with 1% metal film, and the performance will be equal or better than original. Job done!

(Yes, it's remotely possible that a designer would have taken advantage of the poor tempco of a CF to offset something elsewhere, but that's pretty unusual. Bob Pease wrote about that in his book "Troubleshoot analogue circuits")

Most resistors in my stocks are MF, and that's what I use by default. The main exception to that would be where aesthetics are a concern (restoration jobs), and I've had to acquire NOS resistors for these jobs. They won't always be an exact match, but they were available at the time, so it's not impossible to imagine that they could have found their way into the assembly lines.

As a reasonable (but not absolute) rule of thumb, CF are beige and 5%. Sometimes older ones were deep red and 5% (I loved those, BTW!). MF are blue and 1% (often with 5 bands). The BBC like to use MF types from Mullard that were green and 2%. MO are grey, but sometimes blue (with a yellow 4th band). CC are easy to spot because there are no metal end-caps under the coating. Cylindrical axial WW types are often grey and have printed values. Vitreous enamel WW types are dark green. Precision types are different again, but they often have printed values - if so, in common with most printed types, they will have a series number (e.g. RN55C).

That covers most - nearly all - of the resistors in my stores. I say that with some confidence, having re-sorted them all earlier in the year. But the rules might well be different for the US types. Obviously, the service data (if present) helps as well.

Uncertainty when substituting components is the norm. Being able to determine what changes might occur, and then testing for them - as you described back in post #13 - is really all a skilled repairer can do. The more general knowledge of electronic design and component properties one can amass, then the better this process works.

[SteveCG]

Al,

It is many, many years now but I think an IF frequency centred on 150 Mc/s comes dimly to mind for the early 5 Km telescope (the 5 Km was re-engineered over the years).

One little fact I do remember is that the bandwidths of all these IF units were quite wide, yet for the valve IF units used on the (famous for Pulsars) 4 1/2 acre array if a valve was changed then the unit had to be re-aligned - so I suspect stray inductance would have been an issue.

Certainly the era we are discussing for carbon-composition resistor use was when transistors were in use in the IF units, say in the early to mid '70s.

[kalee20]

Originally Posted by mhennessy

MO are grey, but sometimes blue (with a yellow 4th band)..

The Mullard (/ Philips / Vishay depending on what year!) VR25 / VR37 are indeed thus. I had always thought that the yellow band is for 5% tolerance and they didn't want to use metallic gold paint.

[mhennessy]

Ah yes - thank you. Try as I might, I couldn't quite remember the numbers (could only recall "VR").

That led me on to this page: https://archive.org/details/MullardB...nentsMarch1981

I have a hard copy of that at work - the Mullard data books really were great. In there are details of many types of components, and they are well worth reading. The VRxx series are there - and yes, yellow does mean 5% in this context - along with the green MF ones I mentioned earlier (MRxx series). They also mention their beige (well, "tan") carbon film jobs as well (CRxx). Details of HF performance can be found, along with power and reliability data.

Supplementary question: Philips used a lot of film resistors that had a light blue body, but they were rated at 5%. The blue is quite different to the generic MF types, being more of a cyan really. I "recycle" them for my projects (example attached). Are these MF or CF?

[SteveCG]

Al,

A slight correction to my earlier post: The early 5Km telescope definitely had an IF centred on 45 Mc/s. However, as I mentioned, it was re-engineered quite a few times to expand its capabilities, so my recollection of 150Mc/s may come from that.

Still, the electrodynamics, of helical cut resistors has not altered in the intervening years !

[Skywave]

Originally Posted by mhennessy

Regarding your original question (identifying between metal film vs carbon film in an item that you are repairing), I don't really understand why there is cause to worry. Simply replace with 1% metal film, and the performance will be equal or better than original. Job done!

Re the item which I am repairing. FWIW, calling it a 'signal generator' really is an undeserved compliment: it's more of a 'tunable test oscillator': a home-brew job (by someone else) and dates from 1968. Nevertheless, I can think of times when it will be useful. (But let's not go OT over this: this thread is about types of resistors, not overhauling sig. gens., etc.).
And there isn't a "cause to worry" about the type of resistors I used when I overhauled its attenuator: I simply used what was to hand. It was simply that the very doing of that work reminded me that I ought to find out about the various types of resistors that are available these days and their significant properties & visual identification: hence, my OP.
And as for "Simply replace with 1% metal film", that brings us full circle and back to my OP: the visual identification of such types and how to.

Al.

[Skywave]

Originally Posted by mhennessy

It's just a question of being familiar with the families that exist.

Quite: which is the essence of my opening post.

Originally Posted by mhennessy

A quick "Google Images" search for "metal oxide resistors" brings up nearly all the types that I'm familiar with. Same for the other resistor types.

As a reasonable (but not absolute) rule of thumb, CF are beige and 5%; sometimes older ones were deep red and 5%. MF are blue and 1% (often with 5 bands). MF types from Mullard are green and 2%. MO are grey, but sometimes blue (with a yellow 4th band). CC are easy to spot because there are no metal end-caps under the coating. Cylindrical axial WW types are often grey and have printed values.

Those two paragraphs are useful: thanks. I will do a "Google Images" search and add to my growing knowledge of this matter.

Al.

[mhennessy]

Originally Posted by Skywave

And as for "Simply replace with 1% metal film", that brings us full circle and back to my OP: the visual identification of such types and how to.

Not really - if replacing with resistors from your own stores, presumably you know what you have as you procured them from somewhere. The thread has been about correctly identifying what was fitted in a piece of equipment that you are trying to repair - quite a different proposition.

The more I think about this, I'm concluding that the shortest answer to that question is to simply post a good close-up photograph on forums like this one - someone somewhere is bound to recognise it. And it helps others - both forum members, and folk who might be trying the "Google Image" search method.

[Skywave]

Originally Posted by Skywave

And as for "Simply replace with 1% metal film", that brings us full circle and back to my OP: the visual identification of such types and how to.

Originally Posted by mhennessy

Not really - if replacing with resistors from your own stores, presumably you know what you have, as you procured them from somewhere.

. . . . Which assumes that those resistors were obtained from such well-known and reputable sources such as Farnell, RS, CPC, etc. - who would have made available details of those resistors and their properties to the purchaser. But many of us (myself included) will have some components that have been in storage for a very long time - so such records will have disappeared (e.g. from memory, lost paperwork, etc.); some resistors will have been salvaged items; some will have come from 'irregular' sources such as car boot sales, auctions and such like; a private sale from a colleague or friend; a disposal of redundant / old stock from one's employer, etc. From such sources, technical documentation will be very sparse, if any.

Originally Posted by mhennessy

The thread has been about correctly identifying what was fitted in a piece of equipment that you are trying to repair - quite a different proposition.

I don't know how you have reached that conclusion.
Post #1 clearly sets out the objective of this thread. There is no mention of repairing a piece of equipment in that post. My post #3 re-iterates and expands that stated objective of post #1. Again, no reference is made to repairing equipment. My post #10 does make a reference to a signal generator, sure, but it is quite clear that that reference is purely for the purposes of illustrating the 'need to know' about resistor types. Moreover, in post #31, I did state:
"But let's not go OT over this: this thread is about types of resistors, not overhauling sig. gens., etc."

Many replies have been relevant to the stated objective of the first post - and some of those replies have made useful contributions in the areas of RF performance, voltage-rating and noise: the choice of what type of resistor to fit during a repair has not featured in the discussion so far - except for my passing reference to a signal generator, as mentioned above - which no-one else expanded upon. Not wishing to state the blindingly obvious, but resistors are not only used during a repair procedure but also during the design of equipment. And being able to discriminate between the various types and knowing their various properties helps to make an appropriate choice.

Al.

[mhennessy]

Wow! I was merely "reading between the lines" - trying to determine the question behind the question. I took the example given in post #10 as a pretty strong hint. Clearly I was over-thinking

Still, for others reading this thread, particularly those new to the hobby, my advice is sensible: buy wisely and know what you have. Buying new, it's hardly worth considering CF when the superior MF is so economical these days. Second-hand/NOS is mostly fine, but positively identify the type before "releasing" them into your general stores. Many of these types can be very useful - especially when performing sympathetic restorations - but some are best ignored as they are not dependable (e.g. CC types are well known for drifting high with age, and even NOS examples are likely to be well outside of their initial tolerance). Likewise, during a repair, correctly identify the type and understand the significance of any substitutions.

All of which should be relatively easy, thanks to the fine advice from the good people who have responded in this thread. But as I said earlier, if in doubt, a good photo posted on the forum will almost certainly get a result and benefit others in the future