Some random blathering on relays

I wrote the following text a few years ago, with the expectation of posting it, but never did. I'm posting it here in the hope that it will be of some use.

A couple of caveats are in order:

1. Some parts I mention may be NLA ... I wrote this in 2011.
2. The relays, although they can and do fail, are not the weak link. The current handling capacity of the fuseblock wiring and connectors is the big issue here.

With that said, here's what I wrote...

Plug-in relays which are found on the fuseblock are commonly called "mini iso" relays. Their design is specified according to International Standards Organization standards. (ref. 1) I think Bosch pioneered the design, but they are now a common type and are made by Omron, Panasonic, Hella and Tyco Electronics (who bought Bosch's relay business) as well as others. Other styles of relays are now dominant in the automotive world, but these are still very common.

CONTACT MATERIALS

The original Bosch relays I've investigated all have "hard silver" contacts which is a silver/copper alloy. Another common alloy is silver tin oxide (AgSnO2); it is more resistant to welding when making contact to high current loads and might be a better choice when switching lights and motors (which have a high inrush current). Panasonic makes use of this material in their automotive relays. You will also see alloys such as silver cadmium oxide (AgCdO), which has an exemption under the RoHS directive and therefore should still be available (although I think manufacturing of this alloy is going away soon). Given the established lifespan of these parts, I think the contact metallurgy makes little difference in our application, as long as a high quality part is specified.


TERMINALS

The terminals on automotive relays are labeled according to the DIN 72552 standard. Relays have standard nomenclature for the coil and switched contact connections. This is seen on the relays used in Ferrari cars of our era.

Here's a list of the designators and their meanings: (ref. 2)

Relay Coil:
85 relay coil -
86 relay coil +

(Note that Ferrari often disobeys this polarity convention, therefore relays with protective diodes wired in parallel with the coils can't be used without rewiring the circuit the relay is connected to.)

Relay Contacts:
87 common contact
87a normally closed contact
87b normally open contact
88 common contact 2
88a normally closed contact 2
88b normally open contact 2

NOMENCLATURE

In general, relays are described according to the number of switching elements (poles) and the number of contacts they can select from (throws). Thus, the simplest relay which can make or break one contact is Single Pole Single Throw (SPST). Other common arrangements are Double Pole Single Throw, Single Pole Double Throw, 4PDT, etc. This same nomenclature is used for switches.

The notion of "fixed" and "movable" contacts is useful, and I'll mention it here and there. Basically, the movable contact is the one that's doing the switching when the coil is energized, selecting between one or more fixed contacts.

Another notion that is needed to completely describe a relay's switching is the arrangement of the contacts when the coil isn't energized. They could either be open (breaking a circuit) or closed (making a circuit) when the coil is not energized. These are referred to as Normally Open (NO) and Normally Closed (NC), respectively.


RELAY FORMS

The notion of a "form" describes the arrangement of the switching elements more precisely. This is common language in the engineering community, but is not well known in automotive hobby circles. I mention it here as it is helpful to know when doing research. If you look through the manufacturers' literature you'll see things like 2FormA (DPST NO), 2FormC (DPDT), and 3Form. Here's a quick reference:

Form A - normally open contact, single position like SPST, DPST, etc.
Form B - normally closed contact, single position like SPST, DPST, etc.
Form C - double throw or changeover (CO) contact, break before make when switching (i.e. it breaks the old contact connection before establishing the new contact when switching) The familiar Bosch 0332204101 is one of these.
Form D - double throw or changeover (CO) contact, make before break, meaning the two contacts being switched are briefly connected during switching.
Form M - (I have no idea though I've seen it mentioned.)
Form U - This is a DPST relay but with the two movable contacts connected to the same terminal. When it is not energized, the two fixed contacts are not connected to the movable contact nor are they connected to each other. When the coil is energized, they are both connected to the movable contact. You can think of this as a DPDT relay with both movable contacts tied to a single connection. The Bosch 0332015006 is Form U.

References:

(1) ISO7880:1984 "Road vehicles -- Relays -- Arrangement and function attribution of relay terminals" as well as ISO 7588-1:1998 "Road vehicles -- Electrical/electronic switching devices -- Part 1: Relays and flashers" These are the current references AFAIK.

(2) DIN 72552 - Wikipedia, the free encyclopedia

 

n.b. I believe this is mostly correct, but it may contain errors. Doing your own research is recommended.

The following list describes the Bosch relays commonly found in our cars and the replacements I've been able to locate. Note that I've stuck with well-known brands here -- there are many "no name" relays available cheaply, but the slight cost savings seems to be false economy given their unknown pedigree. I've opened up some of the off-brand relays and found bare copper contacts and some dodgy looking construction. I was really surprised by some of the junk I autopsied.

Another note: You will find environmentally sealed relays that fit in the manufacturer's literature. A more detailed read will indicate that a lot of these have reduced current handling capacity and power dissipation specs. This is probably a side effect of their being better insulated and sealed than the standard parts -- so it's a double-edged sword.


Bosch 0332204101

This is a 5 pin SPDT relay (1 Form C) The contact arrangement is Coil 85 & 86, Movable contact 30, Fixed contacts 87 & 87a (30 is connected to 87a with no voltage applied to coil)

Replacement Candidates:

• Tyco V23134A0008X923-EV-590 (from te.com cross reference)
• Tyco (TE Connectivity) / Potter & Brumfield 1432785-1 30/40A, 90 ohm coil, AgSnO contacts, Unplated terminals
• Bosch 0 332 209 137 has resistor, heavy duty version 50A resistive load on NC connection
• Bosch 0 332 209 151 should be suitable replacement.
• Bosch 0 332 209 159 is same as 209151 but with resistor across coil.
• Hella H41388081 (87401) appears suitable, 20/40A contacts, 86 ohm coil, AgNi contacts
• American Zettler AZ9731-1C-12D-C3 (AZ9731-1C-12D-C3E for sealed version) 90 ohm coil, AgSnO2 contacts (From az9731.pdf datasheet)
• Amperite AR2-1C 012D 01 meets all spec, 80 ohm coil, silver alloy contacts, 10 ms max. operate and release times. 100 mΩ max. contact resistance.
• Panasonic CB1-12V (40A Standard type) (I've used these, they work fine)
• Panasonic CB1F-12V (40A Standard type, flux resistant housing)

Zettler (among other manufacturers) offer a series of "sensitive coil" relays with higher coil resistances and higher "must make" voltages. These relays will generally interchange with the parts I've already mentioned and I think that they have both advantages and disadvantages in our application. On the plus side, these have coils that draw less current than a conventional relay. This may be offset by a higher inductive "back EMF". Also, if the voltage available to operate the coil is insufficient, they may be less reliable in this application.

---


Bosch 0332014113



SPST NO relay, contacts 87 and 87a bussed together. (1 Form A2)
Coil 85,86
Movable contact 30
Two fixed contacts, both labeled 87 (internally connected together) Bosch literature calls this a "double 87 terminal"

Resistor across coil to act as suppressor shown in one illustration, but can't confirm from other references.

Replacement Candidates:

• Tyco V23134C0008X901-EV-850 (Cross from Bosch p/n, from te.com cross reference)
• Tyco V23134C0008X901-EV-69A (Cross from internal number, from te.com cross reference)
• Tyco V23134C0008Y901-EV-7Z0 (Cross from internal number, from te.com cross reference)
• Bosch 0 332 019 151 looks like a very good replacement.
• Hella p/n 003510137 (87483) is a suitable replacement (From 2008 Hella Relays catalogue)
• Zettler 9731-1A2-12DC3 (Zettler describes the contact arrangement as Form 1A2, SPNO Single Pole Normally Open [x2])

---

Bosch 0332015006 (1 Form U)

Technically a SPST NO relay, but it has one movable contact and two fixed contacts, presumably to minimize contact wear or increase current handling capability. It can be thought of as a DPST relay with both movable contacts tied to a single connection. This is called "1 Form U"

Coil 85,86
Movable contact 30
Fixed contacts 87, 87b

Rating 32v 2 x 15A (from markings on relay pulled from car)

Replacement Candidates:

• Tyco V23134M0062X915-EV-590 (cross from Bosch p/n, from te.com cross reference)
• Tyco V23134M0062Y915-EV-896 (cross from TE p/n, from te.com cross reference)
• Hella
• Zettler's website does not show a cross from the Bosch part number, but their AZ9731-1U-12DC3 should work as the arrangement of the Form-U contacts shown in the datasheet is the same and ratings are equal or better.
• Amperite AR2-2A 012D 01 should work. Amperite call it a "2 Form A" relay, but the AR2-Series.pdf spec sheet shows "form U" contact arrangement which is the same as the Bosch part. Current rating is higher and all other specs. are the same as the Amperite AR2-1C 012D (listed above).

 

Just a few random notes...

Windscreen Wiper Relay: SIPEA 6.0633.00.0

This is the delay windscreen wiper relay. Autopsies of similar units show a simple timer circuit operating a circuit-board mounted relay which is very similar structurally to what we find inside a mini-ISO relay. I haven't had a failure on the Ferrari, and I haven't determined if these are readily available in the aftermarket. I would be inclined to repair the unit if it failed, or put my own circuitry in it.

If anyone really needs an internal schematic, then perhaps I'll open it up and have a look.

---

Relays under center of dashboard for climate system.

n.b. I no longer have air conditioning in my car. The previous owner started the process and I finished it. So I haven't really explored this area in great detail...

There are three relays in my '83 QV in this location. These are not specified according to part number in any of the literature I have available. Also, the schematic and parts list for these seems to be particularly variable from year to year and model to model.

In my car, the leftmost relay is a 4 pin SPST NC relay in a metal mini-iso package and has no name or other discernible markings other than "12v" stamped in ink. The other two relays are Italamec 4320629 SPST form 1A2. These have black-painted white plastic cases. These Italamec parts seem to be the same as, or at least similar to, the Bosch 0332014113. The leftmost relay is another story...

If the leftmost relay truly has a normally-closed contact arrangement, this might be problematic, as NC relays are not all that common to begin with. Furthermore, while the Bosch 0 332 109 011 is electrically correct, it uses pin 87a for the switched contact (near the center of the relay case) and the original part uses pin 87 (near the edge) for this. So you can't substitute this Bosch part for it, as the contact arrangement is wrong. Similarly, one might be inclined to take a SPDT relay and remove the 87A contact -- while this will mechanically fit, it is the 87A contact that is the "normally closed" one, so this won't work either. (Doing this will basically make the SPDT relay into an ordinary SPST-NO 4 pin relay.)

I will take another look under the dash and try to confirm if this is actually the case.

I have notes somewhere in various methods of protecting relay contacts from erosion when switching loads, as well as controlling back-emf from the coil. If I find these, I'll post them in this thread.

 

When I pulled the relays from my Mondial, I measured their contact resistance, and I also measured the contact resistance of the replacement parts I had selected.

I did this using a HP 3456A meter, performing both a four-wire resistance measurement as well as measuring the voltage drop across the contacts with a current of 1A. I was pretty careful with the protocol, and used the instrument's built-in averaging to improve the quality of the result. My measurement error at these low resistance values was on the order of 10% of the reading based on a couple of standards I could compare them with.

I was surprised by what I found. All of the old relays measured under 2 milliohms contact resistance, with several measuring around 0.1 - 0.2 milliohm. Both measurement methods agreed well with each other. The new relays measured anywhere from 2 to 10 milliohm -- significantly worse but still well within spec.

I think this result is due to the old relay contacts having worn over time to form better mating surfaces. Even units that looked downright ugly on the outside produced great measurements.

Two thoughts:

1. 1 milliohm or 10 milliohm is an inconsequential difference in this application -- the fuseblock wiring and contacts are likely to be orders of magnitude worse.

2. Old relays consistently measure better than new in this case -- suggesting that "shotgun" replacement of all relays could well be a waste of money.

(In spite of this, I still kept the new ones in the car, with the old ones as spares.)

 

Thanks Mike. Yes, it's a definite sleep inducer. My main point in presenting this was to provide a starting point for anyone who really wants to delve into the details of relays, and I was also surprised by the measurement result when I compared new parts to 30 year old parts.

I believe there are ways we can improve the electrical system through proper selection of relays, but I think these gains are small compared to the engineering of the fuseblock and its related connectors. Specifying cheap parts of indifferent design should be avoided.

I think Dave Helms is absolutely on the right track, and I think some of the aftermarket fuseblock designs we've seen here are definite improvements that address the weaknesses of the existing system.

This winter I'll be re-engineering some of that stuff in my car... will post the results.

As always, I think the dictum "Measure, then control" applies here.

 

Randy: Imagine that current is being applied to a coil. This creates a magnetic field (that is storing energy). If the current is then suddenly shut off, the magnetic field created by the current collapses and the energy stored in in has to go somewhere. In what is now an "open circuit" the voltage from the coil will rise until it is able to overcome the circuit resistance. This can cause very high voltage transients to appear in the circuit that was supplying current to the coil. Semiconductors, such as transistors, can be harmed by the appearance of the high voltage. Similarly, a switch used to control the flow of current to the coil may arc, and its contacts can wear from this. Also, the sign of this voltage will be opposite the voltage that is normally found in the circuit, thus producing a negative voltage "below" ground potential, and this can "reverse bias" semiconductors used for switching which can lead to their catastrophic failure.

Placing a resistor across the coil provides a path for that energy to be harmlessly dissipated. Other protective elements can also be used, including a variety of diodes and "snubber networks" made of a capacitor and resistor in series.

This may or may not be warranted in certain circuits in our cars. In general (perhaps all cases?), our relays are controlled by mechanical switches or other relays and not semiconductors. It is also important to note that these sorts of protective circuits can have a deleterious effect on relay life, as they can slow the opening of the relay contacts, which can be destructive. (nothing in life is free).

Edit: Remember when Claude re-wound his fuel door solenoid? We discussed this then as a possible mechanism for eventual switch failure. The only way to know for sure would be to take apart a well-used switch and look for evidence of destructive arcing across the terminals.

A good discussion of this can be found here: http://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=13C3311_AppNote&DocType=CS&DocLang=EN

Jerry: I should read through it and double check my recommendations, just to ensure that it's correct. I thought we had an electrical system thread going at one time... maybe this could go in there? I agree that it would fit well in a parts cross-reference thread.

Anyway, thanks for the kudos. It's just one of those topics that comes up time and again, and I thought we should have some sort of starting off point that answers a lot of the potential questions and provides some background about these things.