• Please review our updated Terms and Rules here

classic Mac analog board 3.9uF capacitor source

coredump4

Member
Joined
May 17, 2019
Messages
31
The compact Mac analog board versions I've encountered all have a 3.9uF non-polarized electrolytic cap, which is somewhat of a unicorn. You can source a 3.9uF film cap from DigiKey (in USA), but it's over US$3. However I found a cheaper alternative: All Electronics has a 4uF, 250V, 105-degree 5% high-frequency film cap for $.50. It's Taitsu brand, which in all fairness I have no experience with, but at least it's Japanese.

(BTW, All Electronics' specs are wrong, it's a 250V unit; see the series datasheet.)

HTH,
Jonathan
 
C1 (3.9µF NP high-frequency horizontal deflection capacitor)

C1 (3.9µF NP high-frequency horizontal deflection capacitor)

WOW! TOO GOOD TO BE TRUE!?

Sounds like you're talking about "C1", the Mac's high-frequency 3.9µF NP horizontal deflection capacitor. Thanks for posting that link; I was just looking for one. At 1/6th the price, that's a great deal! (Hmmm... AllElectronics says I need to add "Only $74.50 more" to my cart to get free shipping. :D )

You wouldn't happen to know a similarly good source of Classic Mac flyback transformers, would you?

VOLTAGE

By the way, you are right about the datasheet. All Electronics is underselling the part by claiming 225V, but even so, that would still be way more than what Classic Macs came with: 25V for the original Mac and 35V for the SE, both at 85°C.

FREQUENCY

My main concern wasn't the voltage but the frequency which All Electronics doesn't even mention. The datasheet has a chart showing max current for all frequencies up to 100kHz, which would be way more than the 22.25kHz frequency it'd be switching as C1. I had been concerned about that because I had been going by advice in Macintosh Repair and Upgrade Secrets which says,

...the physical size of a replacement capacitor is a good indicator of whether or not it's going to work in the C1 application. [...] The tiny 100-V cap on the right is a firecracker waiting to explode. It's completely unsuitable, even though its ratings (4 mfd 100 V) suggest that it might be OK. Use common sense. Don't rely on ratings.

This looks like a teensy-tiny part, so I would have passed this one up. (You already tested this in your Macintosh, right?) Did you see the note in the datasheet that says,

※Please contact us in case pulse cycle exceed 10,000 times

That sounds to me like Taitsu would rather you buy a different part if you're going to be switching faster than 10kHz. (For example, the PMS395J 2E which does appear to be rated for 22.5kHz instead of 10kHz, presuming the "F" column in the chart means "Frequency in kHz").

HEAT


My next concern would have been heat. According to that same book, C1 failed often in hot weather resulting in a single vertical line. The SE's 35V @85°C capacitor could handle the heat better than the 25V part but, interestingly, it is simply because the capacitor is larger so it can dissipate a little more heat. The Taitsu capacitor, being tiny might have some difficulty with that, but maybe it won't matter as much given its much higher working temperature rating of 105°C.
 
Hi,
Interesting points about the frequency. Because the 3.9uF electrolytic is very hard to source, I have used other film capacitors in its place previously. What I've previously used, with success, is the Panasonic ECW-F2395JB. The Taitsu appears to be equivalent to the Pany, but I haven't tried it yet. Looks like a solid choice to me though.
 
Whoa! Careful with that squib!

Whoa! Careful with that squib!

Hi,
What I've previously used, with success, is the Panasonic ECW-F2395JB. The Taitsu appears to be equivalent to the Pany, but I haven't tried it yet.

Um... If you haven't actually tried the Taitsu yet, I'm going to gently suggest that you wait until you contact them as their datasheet says to do and ask them if it can handle 22.25kHz. For comparison, Panasonic's ECW-F2395JB's datasheet specifies it is for computer monitors, so it is actually intended for this application.

Here is the section on replacing C1 in the Macintosh Repair Secrets book which explains why it is so important to get the right capacitor for C1: https://archive.org/details/mac_Macintosh_Repair_Upgrade_Secrets_1990/page/n105/mode/2up

Replacing C1 When choosing a replacement for C1, bear in mind that this is a relatively large size, high-frequency (HF), non-polarized (NP) capacitor. Relatively small size, bipolar (BP) capacitors, the kind used for loudspeaker crossovers and audio filters, won' t work in this circuit. Earlier in this chapter, we learned that the horizontal oscillator frequency (22.25kHz) is 12 times the speed of sound. lf the replacement C1 can' t switch at that speed, its capacitance drops. Instead of passing the signal, it begins to oppose the signal. The charge doesn't clear, the cap heats up, and in most cases, it explodes like a 4th of July firecracker. Not only is that potentially harmful to the set, but it can result in serious personal injury. For safety's sake, don' t experiment with crossover caps. Use one of the replacements listed in Table 5-3.

TABLE 5-3: Acceptable replacement for C1
Cap
WV
WTemp
Type
OEM
3.9 µF100 V85°CHF NP AL radialNichicon
4.0 µF 25 V85°CHF NP AL axialNichicon
3.9 µF250 V85°CHF Metallized polypropyleneNichicon
3.9 µF250 V85°CMetallized polyesterNichicon
3.9 µF100 V85°CMetallized polyesterPanasonic

If none of these are available, the physical size of a replacement capacitor is a good indicator of whether or not it's going to work in the C1 application. For example, the 25-V cap on the left of Figure 5-5 is an OEM part removed from a stock Mac Plus. The large 100-V cap in the middle is a better-than-OEM replacement. The tiny 100-V cap on the right is a firecracker waiting to explode. It's completely unsuitable, even though its ratings (4 mfd, 100 V) suggest that it might be OK. Use common sense. Don't rely on ratings!

Figure 5-5
squib.jpg
PhysicaI size is a good indicator of whether or not the replacement capacitor is going to work. The little nonpolar cap on the right is completely unsuitable, even though its ratings (4 mfd 100 V) suggest that it might be OK.

I know size isn't everything, but the Taitsu part is only a quarter of the volume of the Panasonic part (3,764 mm³ versus 16,607 mm³), so I'd be wary of it.
 
Last edited:
(For example, the PMS395J 2E which does appear to be rated for 22.5kHz instead of 10kHz, presuming the "F" column in the chart means "Frequency in kHz").

Minor correction to my previous post: the "F" column is a measurement of length of the part in mm, so nothing to do with frequency. The PMS395J 2E is also labeled by Taitsu as “※Please contact us” if you are going over 10kHz.
 
Hmmm, it still looks like a legit replacement to me, but you're right about being cautious; I'll e-mail them.
 
Well, sorta. I asked if their PMJ caps would work in a horizontal oscillator circuit at 22.25kHz. They replied that the "general advice for S-shaped correction circuit is PMS or PMJ, and for horizontal resonant circuit is PHS type." But their PHS line has the same warning as the PMJ, so I'm still unsure.

It doesn't make sense to me that your marketing literature would identify the caps as "high-frequency," but they wouldn't work at a few kHz above audible frequencies. I guess that isn't what I think of when I see "high-frequency."
 
I sent a follow-up e-mail to them asking for clarification re: using PMJ, PMS or PHS above 10 kHz.
 
Huh? :huh:

Someone want to 'splain that one to me?

Oh, it was a silly thing where the author calculates out the rate of the raster across the screen. Despite the ridiculousness of that, I do take his warnings about incorrect flyback capacitors exploding seriously.
 
I sent a follow-up e-mail to them asking for clarification re: using PMJ, PMS or PHS above 10 kHz.

I followed up with Taitsu and asked for clarification about the cryptic warning in the datasheets. Their response was:
Code:
If the waveform used is one that does not care about pulse current, you do not need to care about pulse cycle.
Does that make more sense to all of you than it does me? :p
 
It sounds to me like they're assuming that you'll use this cap in a switching power supply, not a CRT application. They're talking about pulse width vs. pulse current for charging the bulk output capacitors in a PSU, I think.

That's at least what I could glean from that.
 
FYI I used one of the above Taitsu caps in the horizontal section of a 9" Apple CRT (forget the model, but it's the monitor for the IIc) to replace a 4uF NP electrolytic there. It worked fine to repair the monitor and we'll see about longevity I suppose.
 
Back
Top