Heathkit Model CP 1060 Capacitive Discharge Ignition

[John1397]

One of my transistor has failed I tried replacing but sticking any transistor in there does not seem to work. Someone from that worked for Heathkit told just look up number this would work except I can not find cross for 230-616. When this unit worked it always put a little wine out I think it was from the transistors but not sure. Need to know more info like base to collector voltage and power rating.

 

[CWatters]

Is there anything written on the transistors themselves?

 

[Asymptotic]

Do the transistors look like this? Photo is from a Mark 10B, but the transistors look identical to ones pictured in an Archerkit.

From what I've read, the Heathkit CP-1060 kit was substantially identical to a contemporaneous Radio-Shack Archerkit, and Delta Products, Inc. of Grand Junction, Colorado was the source for both. I haven't learned much about the transistors except were made specifically for Delta (note the Δ sigil). For what it is worth, "230-616" puts me in mind of Radio-Shack nomenclature.

One of my transistor has failed I tried replacing but sticking any transistor in there does not seem to work.

The "sticking any transistor in" method of parts substitution worries me. What makes you sure the transistor has failed?

Is this CDI used in a negative ground, or in a positive ground system? The schematic shows PNP transistors, but they well could be NPN.

A Delta Mark 10B featured in an advertisement from Popular Electronics, Jan 1973.

References
https://www.physicsforums.com/threads/how-do-capacitor-discharge-ignition-circuits-work.384566/
http://selectric.org/delta/index.html
https://www.bmw2002faq.com/forums/topic/136076-delta-mark-ten-cdiradioshack-archerkit-question/
http://www.alfabb.com/bb/forums/electrical-chassis-lighting/174312-delta-mark-ten-b-ignition-box.html
https://e9coupe.com/forum/threads/electronic-transistor-ignition-useful.10375/

 

[jim hardy]

Can't see your whole schematic.
Hopefully there's nothing too special about those transistors. Only potential trouble spot i see is Vbe rating .

do those two transistors get power from 12 volts ? Probably.

A circuit like that is usually not too choosy about transistor parameters, it self-oscillates by positive feedback so it'll tolerate quite a bit of slop.
The transformer's time to saturate determines frequency.
The transistors are just switches. When one of them turns on, positive feedback via tapped transformer winding reinforces its base drive. Same feedback in other half of winding opposes the second transistor's base drive turning it hard off.
That condition persists until the transformer saturates which removes 'base boost' from the transistor that's conducting and removes 'base holdoff' from the other, so the transistors swap states.
i'd think a fresh pair of 2N2955's would likely do the job. They're TO-3 case like in @Asymptotic 's photo.
http://www.unisonic.com.tw/datasheet/2N2955.pdf
If yours are TO-220 case you're in luck there are lots more power transistors to choose from.

Back in the days of Heathkit , high power PNP's were sort of scarce.
2N 2955 has decent gain and pretty robust voltage withstand.
I can't quite make out the resistor values in your image. If you need more gain you could look into power darlingtons and adjust the bias resistors..


Don't forget to put heat transfer grease under them. Dow 340 is made for that but Dow 4 from the hardware store will work.

my two cents, and overpriced at that...

old jim

EDIT found a higher current transistor , mje5025, but Vbe is 5 volts vs 7 for 2N2955..
https://www.onsemi.com/pub/Collateral/MJ15023-D.PDF

 

[John1397]

That's the way they looked and them number's wouldn't cross either I kind of figured they where made for this specific application If I remember I just used a power amp transistor. I tested both and I found one was good putting just one in that is different from the other is not good. It was new car that used negative ground and it worked just as stated points never burned just wore off rubbing block.

 

[jim hardy]

...them number's wouldn't cross either I

What numbers wouldn't cross?
2N2955 is industry standard part number , they're all over Ebay

MJE5025 is popular with audio guys https://www.parts-express.com/mj15025-transistor--mj15025

Another candidate is 2N3792, also readily available
datasheet : http://www.mouser.com/ds/2/523/2N3791-601092.pdf
60 in stock here: https://www.mouser.com/ProductDetail/Microsemi/2N3791/?qs=TXMzd3F6Eym8M0G3OdytqA%3d%3d&gclid=EAIaIQobChMIjbnEjNCF1wIVDKtpCh2XZAGfEAAYASAAEgIQ5fD_BwE

 

[Asymptotic]

That's the way they looked and them number's wouldn't cross either I kind of figured they where made for this specific application If I remember I just used a power amp transistor. I tested both and I found one was good putting just one in that is different from the other is not good. It was new car that used negative ground and it worked just as stated points never burned just wore off rubbing block.

Does your meter have a 'diode test' function? What are the collector-base and emitter-base forward voltage drops on the good transistor?
They will be about 0.7V for silicon transistors, and 0.3V for germanium.

Delta Products sold their Mark Ten either assembled or as a kit, and although it wasn't uncommon for circuit developers of that era to write an article for Popular Electronics or Radio-Electronics featuring their kit I couldn't find one was written for this particular circuit. However, mention of the Mark 10 is made as early as 1964, and parts list for other CDI circuits from this time showed germanium power transistors like the 2N173 and 2N1100 being used. These were in TO-36 cases, not TO-3, but TO-3 case germanium power transistors were around by then, and the Mark Ten could have used something along the lines of a 2N1532.

The June 1964 issue of Popular Electronics has an excellent article on early transistor ignition systems including a basic diagram of the Mark Ten circuit. A slightly different CDI circuit project using 2N1100 transistors was featured in the June 1965 issue, and is also well worth reading.

Scanned PDFs of Popular Electronics, Radio-Electronics, and other such tech magazines are available at the American Radio History website. Here is a link to the Popular Electronics index.

 

[jim hardy]

They will be about 0.7V for silicon transistors, and 0.3V for germanium.

Great check to make.

I can't make out values
maybe @John1397 could use Paint to annotate part of his drawing

a 15 and a 430 would bias a germanium barely into conduction, 0.40 volts
a 25 and a 430 would bias a silicon barely into conduction, 0.66 volts

to assure it'll self start.

 

[jim hardy]

Hmmm Good point on germanium , @Asymptotic , i tend to forget those days.
i had forgot Germaniums had such high emitter-base reverse breakdown

http://www.datasheets360.com/pdf/-419200064720405383
might have to protect silicons against that unless we know what voltage is between those transformer taps..

 

[John1397]

Must have been geranium I remember testing the old one wonder if it used geranium if one could use silicon?

 

[jim hardy]

wonder if it used geranium if one could use silicon?

We don't know if there's enough voltage in the 'tickler' section of that winding to over-reverse-voltage the EB junction.
If you're equipped to remove both transistors and apply 12 VAC across orange to red taps, then measure voltage between ends you'll have that information.
That'd give a feel for whether to choose a E-B parallel zener or a series blocking diode with base for EB protection.
A 1N5913 3.3 volt 3 watt Zener connected EB should shunt-protect a silicon transistor's EB junction.
Mouser says they have several thousand in stock, around $0.38 each in small quantity
http://www.onsemi.com/pub/Collateral/1N5913B-D.PDF
but feel it and the low ohm resistor to make sure neither gets overly hot.


Did you figure out what value are those biasing resistors in original configuration ?

 

[Asymptotic]

Did you figure out what value are those biasing resistors in original configuration ?

Excerpted from a CP-1060 manual posted to radiomuseum.org.

15Ω, 1 watt resistors for b-e on both transistors. Q1 has 1KΩ from c-b while Q2 uses 470Ω. Donnerwetter! ... Germanium is looking more and more likely.

Hmmm Good point on germanium , @Asymptotic , i tend to forget those days.

It came to mind while reminiscing on a friend's band, and an early '60s PA amp head that ate germanium power transistors like candy. Totally unrelated to the subject at hand, but germanium signal transistors are still favored in "vintage sounding" guitar effect boxes, and the tone Brian May achieved playing several of his iconic guitar pieces came from 'Deacy' - a germanium audio amp salvaged from a dumpster.

 

[jim hardy]

Germanium is looking more and more likely.

Yep. Looks like Q2 gets to do the kickstart.
Old transistors. Wow. How soon i forget - yet memories of high school electronics class seem nearer than this morning's breakfast 2N1532 has suggested replacement NTE121 at about $35 apiece most places i looked.
http://www.farnell.com/datasheets/1693972.pdf
Found a place in Ohio offering surplus 2N1532's at $ 5 apiece
I've never dealt with them so cannot vouch for them. Look over their site for yourself. .
They claim to have been around since 1946 which is comforting. Still, I'd call and feel them out before ordering.
http://www.electronicsurplus.com/amc-2n1532-transistor-pnp-p-n-2n1532
Get some spares .

 

[jim hardy]

RE tickler winding voltage :
1 watt in 15 ohms would be about ¼ amp
well within 1N5913's 454 ma current capability
looks to me like shunt EB protection by zener just might just work okay.

that is if he wants to change to silicon.
Caveat - Some trial and error may be required to arrive at biasing resistor values because things seldom go as easy as they look like they should. We all know Murphy was an incurable optimist.

 

[jim hardy]

Another reason to change to silicon

http://ieeexplore.ieee.org/abstract/document/4052039/

 

[John1397]

I did not look up any specs but I my box has a 2N3791 transistor in there and this did not make it work.

 

[jim hardy]

I did not look up any specs but I my box has a 2N3791 transistor in there and this did not make it work.

What does it do ? anything getting hot ?

Specs are at
data http://www.mouser.com/ds/2/68/2n3789-3792-42519.pdf

The R1-R2 and R3-R4 voltage dividers don't make enough voltage to turn on a silicon transistor.
You'll have to adjust the bias resistors to get about 3/4 volt drop across the 15 ohm ..
If your 3791 is Q2 try tacking 250 to 300 ohms in parallel with R3.
If your 3791 is Q1 try 200 to 250 in parallel with R1.

Place a 1N4001 between base and emitter of your 3791, cathode to emitter and anode to base.

 

[John1397]

I am going to try and get a geranium it got 100000 miles of use before it quit working.

 

[Asymptotic]

Tweaking the circuit and changing over to silicon transistors as @jim hardy suggested is definitely the way to go. Most of the possible germanium candidates found on eBay, Amazon, industrial parts sources such as Radwell, and elsewhere cost and arm and a leg.

Also, replace 'C1' as a matter of course. Good money says a 40+ year old, 400 uF electrolytic in this application has either failed, become titantically leaky, or has both feet firmly planted on banana peels.

Additional research cites
A comment in this Softsolder article suggests power transistors in the original Mark Ten (apparently, they had been TO-36 early on, then changed to TO-3 in later production runs) may have been 2N1523. From alltransistors website:

Maximum Collector Power Dissipation (Pc): 75 W
Maximum Collector-Base Voltage |Vcb|: 80 V
Maximum Collector-Emitter Voltage |Vce|: 60 V
Maximum Emitter-Base Voltage |Veb|: 30 V
Maximum Collector Current |Ic max|: 50 A
Max. Operating Junction Temperature (Tj): 85 °C
Transition Frequency (ft): 0.04 MHz
Forward Current Transfer Ratio (hFE), MIN: 25
TO-36 case

The 50A Ic spec seems high, but vast derating may be one of the reasons germanium power transistors held up so well here.

Mark Ten CDI modules come up in an article at speedtalk.

... The NTE 5438 is a good replacement for the SCR, as you mentioned earlier. You will have to bend the leads sideways and cut the leads to fit like the original SCR, touching against the case. Use a small needle nose pliers or equivalent when soldering for a heat sink and bending the leads so as not to damage the SCR case. The replacement SCR also needs to be the same height so it hits the housing bottom, as the bottom serves as a heat sink for the SCR. Buy a couple of them while your at it. Some heat sink grease is not a bad idea.

The Delta Mark 10's use a Germanium PNP transistor with the electrical equivalents similar the 2N3055. An SK 3009 should work, if you can find one.
...
American Capacitor Corp. also sells a good capacitor if you feel the need to replace it. # VW2J105K is a 1.0 MFD, 600 volt cap. The 1.5 MFD Cap. #VW2J155K 600 volt is probably the capacitance the Mark 10B came with. I usually replace that with a 1.0, as the smaller cap seems to pull revs better. As clshore said, these will almost knock on your butt if you allow them to. I strongly recommend not pushing down on the wires in the dist cap to make sure all the wires are in all the way if the engine is running. You may find that the wire boot is not as thick as you thought it was.

Also, the Delta's came with both size caps, so you can use what you want and not be wrong. Make sure you feed it a full 12 volts, and bypass any resistors in the coil feed wiring.

 

[jim hardy]

I am going to try and get a geranium it got 100000 miles of use before it quit working.

From @Asymptotic's link,

The Delta Mark 10's use a Germanium PNP transistor with the electrical equivalents similar the 2N3055. An SK 3009 should work, if you can find one.

Hmmm... same surplus place in Ohio says he has a few SK3009's for ten bucks apiece.

http://www.electronicsurplus.com/rca-sk3009-transistor-pnp-p-n-sk3009

In my years as a troubleshooter I've seen germanium transistors degrade with age - they lose gain, get 'leaky' and temperature sensitive.
If you can , mount this thing in front of the radiator where it's so much cooler than engine compartment.

If you do stick with germanium i'd say buy a few spares..

 

[Asymptotic]

I've been having a fascinating time reading old magazines on the American Radio History site, and in the process sussed it out in the November, 1964 issue of Electronics World, "SCR Automotive Ignition System" by Brice Ward. It is somewhat different than the later version used in Heathkit CP-1060 (no 400 uF cap, no RF choke, no SCR gate resistor, etc.) but they are clearly related.

Inverter transistors are listed as 2N456A, and while I'm betting another difference between the original kit circuit was upping transistor ratings, the 2N456A gives a firm starting point.

2N456A crosses to an NTE104 (still available from various vendors at about $25 a pop). I haven't bought anything from them in decades so can't say how they are now, but http://www.bgmicro.com/pnp-germanium-house-number-to-3-like-2n176-2n256-2n555-nte104-and-nte121.aspx claims to have equivalent parts for $1.56/each; cheap enough to cause no heart break if they don't work out.

 

[jim hardy]

For us Old Guys - 1983 RCA Solid State Applications Manual available for download here - 288 pages of nostalgia
https://ia600900.us.archive.org/22/items/RcaPowerTransistorApplicationsManual/RcaPowerTransistorApplicationsManual.pdf

saved a copy in my FOR PF folder.

 

[Asymptotic]

For us Old Guys - 1983 RCA Solid State Applications Manual available for download here - 288 pages of nostalgia
https://ia600900.us.archive.org/22/items/RcaPowerTransistorApplicationsManual/RcaPowerTransistorApplicationsManual.pdf

saved a copy in my FOR PF folder.

Thanks for the link!

In my opinion, the value of vacuum tube, transistor, and other device manuals from yesteryear extend far beyond nostalgia alone. Their stripped down, "this is what you absolutely need to know" presentation style appeals to me, and although certain examples don't have much practical application these days (color TV chroma, etc.) the majority of circuit examples remain valid.

Background information (basic theory, protection circuits, thermal considerations, and other 'best practices') haven't changed much, if at all.

This RCA manual has a nice section from page 243 to 256 on transistors in automotive design, including CDI.

 

[jim hardy]

"this is what you absolutely need to know" presentation style appeals to me, and although certain examples don't have much practical application these days (color TV chroma, etc.) the majority of circuit examples remain valid.

Discrete transistors in electronics are about like rivets in shipbuilding - almost obsolete. Still, they teach one circuit skills.

 

[John1397]

I would think all modern engines would have this same setup to fire the plugs.

 

[Asymptotic]

For us Old Guys - 1983 RCA Solid State Applications Manual available for download here - 288 pages of nostalgia
https://ia600900.us.archive.org/22/items/RcaPowerTransistorApplicationsManual/RcaPowerTransistorApplicationsManual.pdf

saved a copy in my FOR PF folder.

"RCA Solid-state Power Circuits Designer's Handbook" #SP-52 is mentioned multiple times in the power transistor handbook. It is available in the Internet Archives open library, and has a wealth of detail information on designing power inverters including transformer core steel selection (pages 302 to 338) and specifically on capacitive ignition systems (pages 678 to 692).

 

[Asymptotic]

I would think all modern engines would have this same setup to fire the plugs.

Mechanical distributors have long been disappearing from modern ignition systems in the direction of a "coil on plug" approach. I don't keep up, but it seems several different variations on this theme are used. Coil-on-plug Ignition (AA1Car article).

 

[wcubed]

I built and repaired several of these in the 70s in high school. The capacitors were all high quality Siemens „block“ style. I never liked the way the SCR Mount worked with the brass spring heat conductor to the bottom of the case, which was lost very easily. The Power Transistors didn't fail as much as the SCR did. My electronics teacher sold one to almost every teacher on campus. With the high failure rate, the „b“ model with the switch was handy.