开云体育

Hello,
The recent discussion of various 22xx fan power supplies reminded me of an odd fan supply I saw awhile back, but I couldn't remember on what 22xx model it was.
Some searching turned up this odd fan circuit on the high serial number 2236 schematics which I scanned and will put up on Leo's photo album, as it might be interesting for others to see.
I mentioned in the previous discussion how the 2236 had a half wave fan supply, but that only turned out to be true for the lower serial number range of the 2236's.
At S/N B022267 Tektronix changed the fan power supply to an odd combination that is both full wave and half wave. I don't know much about electronics so this all seems odd to me.
One leg of the fan is supplied by a half wave negative supply from pin 20 of the 8.6 volt supply with a diode,? a 2.7 Ohm resistor and a 4700 uF capacitor going to ground.
The other leg is supplied directly from the full wave rectified positive voltage coming from pins 17 and 19 of the 5 volt supply to the scope.
It seems like every one of these 22xx fan power supplies is a bit different, yet they still only use three components C965, CR965 and R965 in most or all(?) of them.
Of course the component values vary widely from model to model.
As an example on these two 2236 fan power supplies, the capacitor C965 is 270 uF on one and 4700 uF on the other, and R965 goes from 20 Ohms on one to 2.7 ohms on the other.
Take a look at these two 2236 fan schematics if you are interested, they are in Leo's 2215A photo album at:

/g/TekScopes/album?id=64919

and please post any thoughts you have about them.
tom jobe...

PS The serial number at which the fan power supply changed is a guess on my part, as I used the high serial number electronic parts listings to see when C965 changed from 4700 uF to 270 uF.

Thanks
As you can see I a not a chemist.
I will order some 95% formic acid soon.

Glenn

--
-----------------------------------------------------------------------
Glenn Little ARRL Technical Specialist QCWA LM 28417
Amateur Callsign: WB4UIV wb4uiv@... AMSAT LM 2178
QTH: Goose Creek, SC USA (EM92xx) USSVI LM NRA LM SBE ARRL TAPR
"It is not the class of license that the Amateur holds but the class
of the Amateur that holds the license"

Thanks
As you can see I a not a chemist.
I will order some 95% formic acid soon.

Glenn

--
-----------------------------------------------------------------------
Glenn Little ARRL Technical Specialist QCWA LM 28417
Amateur Callsign: WB4UIV wb4uiv@... AMSAT LM 2178
QTH: Goose Creek, SC USA (EM92xx) USSVI LM NRA LM SBE ARRL TAPR
"It is not the class of license that the Amateur holds but the class
of the Amateur that holds the license"

You don't necessarily have to have an air connection to the outside world for a fan. You can get some beneficial cooling by circulating the air inside the instrument, spreading the heat for dissipation through the walls. This would only be worthwhile if there are certain hot spots that are troublesome - if you can direct some air from an internal fan, it will cool the target area, and increase the average temperature in the rest.

For those contemplating adding a fan to 2200 series scopes, I'd recommend using full-wave rectification on whatever windings are intended for this. Even if there are only built-in board features for half-wave, it's best to just add the extra diode somehow. Full-wave will be "nicer," and better balanced, so less likely to increase the PS noise.

The HF PS signals on the windings are more or less square, with nicely rounded edges, so the filter capacitor could be quite small, in theory - but that's only for a constant DC load. The reason the cap needs to be fairly large, considering the switching frequency, is to filter and isolate the fan commutation ripple current, so it doesn't reflect into the PS. Depending on the particular fan motor characteristics, you may need more or less C.

Ed

Just realized... stupid me ! ..... There is NO WAY one could fit a fan in a 2215 if like mine (and most units AIUI), it has the modified SMPS which replaced the TRIAC based regulator with the PWM / FET daughter board ! The board is mounted upside down and some of its components lie where the fan would go, most notably the big 400V mains filter cap, to start with !

Oh well....


Vincent Trouilliez

On Sun, Aug 5, 2018 at 08:22 AM, tom jobe wrote:

As odd as it might sound, the 2215 is very unlike the 2215A (in many ways) and
what is known about adding a simple cooling fan does not apply to the 2213 and
2215 as well as many of the other small 22xx scopes that were made around the
world.
Even common scopes like the 2213A do not have this undocumented fan circuitry
on the mainboard.

Hmmm.... yes indeed !

Just looked at my 2215... no luck, as you said... :-/
At first glance I thought I stood a chance when I spotted C965 and CR965 (but no R965 and no fan connector) on the board layout in the manual... but as it happens they have nothing to do with the fan circuitry... they are used to rectify and filter the +30V rail, bummer...

Well, the grill / vent for the fan is there none the less, so from a mechanical point of view, one could still mount a fan in there.
From an electrical point of view, we have said plenty enough in this discussion to be able to implement the required circuitry and how to wire it to the secondaries. Only a few components to mount an a tiny bit of vero board, secure it somehow in the SMPS somewhere, with adequate insulation... and you are good to go. :-)

I won't bother doing it on my scope as it's not needed... it's only a backup scope and if I ever use it, it would not be powered for any length of time... I don't let any of my gear powered up all day long, never mind over night. Plus, I have spent enough money on this budget scope already... not going to add a fan to the list.

Would be different if it were my main scope, of course... but my main scope is a 2232, and it has a fan already.


Vincent Trouilliez

On Sun, 5 Aug 2018, Glenn Little wrote:

I use acetic acid (vinegar) to clean alkaline battery residue and baking soda in solution to clean acid battery residue.

Formic acid is better than acetic -- it is more volatile than water or IPA
so whatever left after the cleaning will simply evaporates without a trace.
Don't use concentrated formic acid -- it has extremely strong odor (kinda
like acetic but many times stronger), it burns skin and will etch too much.
Dilute it with something like 5 parts of water.

As for baking soda for cleaning acid -- it is also not the best choice.
Ammonia is much better for the same reason as formic acid -- ammonium salts
are volatile unlike sodium. And excess ammonia will simply evaporate by
itself. Sodium sulfate, on the other hand, is difficult to remove completely
and it will keep etching your board. It is a very slow process but sooner or
later (couple of years) it WILL etch holes in copper traces.

---
*
* KSI@home KOI8 Net < > The impossible we do immediately. *
* Las Vegas NV, USA < > Miracles require 24-hour notice. *
*

Okay, I've found some additional evidence of what's gone wrong. Both R1300 and R1301 have exploded. I'm curious what I should try replacing and how I should turn the thing back on after the pieces have been replaced. I don't want these things to explode in my face if something else is causing them to fail.

I use acetic acid (vinegar) to clean alkaline battery residue and baking soda in solution to clean acid battery residue.
I have a wash bottle of both on the work bench.
I follow this up by flooding the area with distilled water with the board angled such that the contaminated area is closest to the bottom.
This is followed by cleaning the area with IPA then warming with a hot air gun.

Ensure that you neutralize any residue that has made its way into vias.
If the board is multi layer and the residue has gotten into the inner layers, the board may take heroic efforts to salvage it.

Glenn

--
-----------------------------------------------------------------------
Glenn Little ARRL Technical Specialist QCWA LM 28417
Amateur Callsign: WB4UIV wb4uiv@... AMSAT LM 2178
QTH: Goose Creek, SC USA (EM92xx) USSVI LM NRA LM SBE ARRL TAPR
"It is not the class of license that the Amateur holds but the class
of the Amateur that holds the license"

I use acetic acid (vinegar) to clean alkaline battery residue and baking soda in solution to clean acid battery residue.
I have a wash bottle of both on the work bench.
I follow this up by flooding the area with distilled water with the board angled such that the contaminated area is closest to the bottom.
This is followed by cleaning the area with IPA then warming with a hot air gun.

Ensure that you neutralize any residue that has made its way into vias.
If the board is multi layer and the residue has gotten into the inner layers, the board may take heroic efforts to salvage it.

Glenn

--
-----------------------------------------------------------------------
Glenn Little ARRL Technical Specialist QCWA LM 28417
Amateur Callsign: WB4UIV wb4uiv@... AMSAT LM 2178
QTH: Goose Creek, SC USA (EM92xx) USSVI LM NRA LM SBE ARRL TAPR
"It is not the class of license that the Amateur holds but the class
of the Amateur that holds the license"

Did someone here suggest Ammonia (Ammonium Hydroxide solution, Washing Ammonia) as a way of removing the crud from failed alkaline battery corrosion?

Dick

Hi Roy,
As odd as it might sound, the 2215 is very unlike the 2215A (in many ways) and what is known about adding a simple cooling fan does not apply to the 2213 and 2215 as well as many of the other small 22xx scopes that were made around the world.
Even common scopes like the 2213A do not have this undocumented fan circuitry on the mainboard.
The common 22xx scopes this fan modification applies to are the 2215A, 2235 and the Military version of the 2235, the AN/USM-488 and probably the 2235A as well.
The component numbers are printed on the mainboard of the 2215A and the 2235, but the markings are gone on the military version even though the unused fan circuitry is there.
Higher models such as 2236, 2230, 2232, 2221 come with a fan and a fan circuit that is similar. Some are half wave like the 2236, others are full wave like the 2232.
Of the ones I have seen, they all use pin 20 on the main transformer, and also use pin 22 if it is a full wave supply on that particular model.
Below is a document from 2008 that explains the fan circuit and components I used, and includes some of the good input I got from other Tekscope members back then.
(Unfortunately the photo album capability on this group will not allow posting this related document with the photos because it is not an 'image' file)
tom jobe...

PS I have installed the fan circuit on quite a few 22xx scopes in the last 10 years, and the uF value of the capacitor does not seem to matter much, nor does going to a full wave supply make much difference in the end result. Adding another diode over to pin 22 of the transformer, on the bottom side of the mainboard is all you need to get a full wave supply if you want to try it, and a photo of that will also be added to Leo's photo album after I edit the file names to help explain what is in each photo. I see someone else has posted some photos on this same subject to Leo's album but with no hint of what you are seeing in the file names.


Components I have used include:

CR965 = ultra fast diode, one amp

R965 = 3 ohm resistor, 1/4 Watt

C965 = 100uF to 1000uF capacitor

Recently the question was asked "was there an optional fan for the AN/USM-488 oscilloscope (the military version of the 2235)? and some off line discussion led to some interesting findings that a few of you might like to hear about.

There was no optional fan as far as anyone seems to know, but there are some unused traces and solder pads on the mainboards of the 2215A, 2235, and USM-488 (and other similar 22XX's), that appear to have been made for a power source to run such a fan.

On the 2215A and 2235 I was looking at, the component numbers were silk screened on the mainboard in these empty locations, and these component numbers are not in the electrical parts listings or shown on the schematics I looked at. The USM-488 does not have the markings on the mainboard, but all of the same traces and solder pads are there.

Looking at the front of the oscilloscope, these component locations are near the right edge of the mainboard, just in front of the aluminum plate that is the heat sink for the two inverter transistors and the one MOSFET of the inverter.

It is a half wave supply coming from pin 20 (the 8.6 volt winding) of the main transformer to the cathode of diode CR965. The anode of CR965 connects to the resistor R965 which becomes the minus pin of the 2 pins called P9965 (which is where the fan plugs in). The positive pin of P9965 is connected to ground and a capacitor C965 goes between the plus and minus pins of P9965 to smooth the half wave power. (the fan's positive supply is ground, and the fan's negative supply comes from this negative supply circuitry)

I did not know what to do for component values so I studied the full wave supply for the 2232 fan and my electronic text books for awhile, and came up with these values. CR965 is a one amp diode, R965 is a 3 Ohm resistor, and C965 is a 1000 uF capacitor.

The resistor limits the current to protect the diode at start up, and the capacitor was probably intended to be fairly large as the capacitor mounting holes are on 0.3" centers (7.6 mm). The capacitor for the 2232 full wave fan supply is a 4700 uF, so 1000 uF seemed like a reasonable place to start, and it physically fit in there nicely.

The fan is a 60 mm / 12 volt computer style fan, and it fastens to the inside of the back of the metal case where the air holes and the square 50 mm hole pattern is. All of the 60 mm fans seem to have the 50 mm bolt pattern, and I happened to find one fan that had threaded inserts in the four holes so no nuts were even required. The fan is on the inside of the case blowing the air outwards just like your computer does.

I did this fan experiment on a 2215A I have been working on for some time. With this modified oscilloscope running, the fan is getting about -8.4 volts, which keeps it fairly quiet yet it does move some air, and there is about 60 mV of ripple at the fan connection.

Next I thought I would try making this a full wave fan supply, so I added another identical diode and a piece of wire (all in shrink tubing), to the back of the mainboard. This second diode's cathode connected to pin 22 of the main transformer (the other end of the 8.6 volt winding) and went over to the end of resistor R965 where the first diode CR965 connects. The full wave arrangement delivers about -8.6 volts (up from -8.4) and the same 60 mV ripple so it seemed foolish to bother with the full wave configuration and I removed it. In one of my books it talked about how you need about 0.5 volts of ripple in the first stage of a power supply to make sure both diodes are actually working, and I only had 60 mV.

tom jobe

…. tom jobe wote

Hello John,

Your mathematical explanation helps makes more sense of this for me.

This morning I was reading in one of my older text books about all of this,

and they gave a rough formula to figure out the approximate capacitor

required for a given ripple on a half wave rectified 60 Hz sine wave

circuit.

C = It/V

where "C" is the capacitor size in Farads, "I" is the load in amps, "t" is

the time in seconds, and "V" is the allowed ripple in volts.

If you plug in .12 for the fan load in amps, times .000050 for the 50 uS

period of 20 kHz, and divide by .05 for 50 mV of ripple, it shows you that

you only need 120 uF

The square wave would have a higher RMS value than the sine wave, so I guess

the capacitor could probably be even less, maybe only 100 uF?... if I

haven't messed this calculation up somewhere!

If there is no real downside to leaving the 1000 uF capacitor in there? I

think I will just leave it alone, and move on.

I've had a good time with this project, and I thank you for your help!

tom jobe...

tom jobe wrote:

Hello John,

Thank you for the additional insight into the capacitor sizing. I'm in

way

over my head, so I can use all of the help I can get to understand this!

One thing I failed to mention was that this transformer runs at 20 kHz,

and

the winding you are connecting the fan to is an approximately 19 volt

Peak to peak square wave with a period of 50 uS. (this square wave meters

18.4

VAC with an RMS DVM)

OK. The same formula applies, but the dT is 25 or 50 uS. With a square,

the cap

will be oversized.

The square wave is fairly symmetrical, and when it was being rectified

full wave, the resulting trace (without a capacitor) was not a perfect

straight line of course, but it wasn't bad. The full wave rectified square wave

trace is made up of approximately straight pieces on the tops and bottoms of

the square waves with little glitches in-between every 25 uS, so the smoothing

probably does not require nearly as much capacitor as a full wave rectified sine wave, or half wave rectified square wave would.

Correct.

I was surprised at how the ripple did not change much when the fan

Supply was changed back to half wave rectification, but I guess the fan uses so little power (0.12 amp?) that the capacitor does not get pulled down very much in the 25 uS between half wave pulses.

The ripple depends on the load current and the cap size. With a square wave, current is supplied

throughout the cycle, hence the lower ripple.

Arthur pointed out in another reply to this subject, that 1000 uF is

probably far too much capacitor for this little fan supply. What made me

start with 1000 uF was two things, one was that a study of the similar

full

wave fan supply in a 2232 has 4700 uF (was Tektronix trying to tell me

something?), and the other thing was the mounting holes for the

capacitor

were on 0.3 inch centers so the smaller capacitors would not sit down on

the

board properly because their lead centers are closer together at 0.2

inch or

less. Also. I could not see that it would hurt anything to have a larger

capacitor if the current was limited through the diode at start up, even

though I am questioning my choice of resistor size at just 3 ohm. (the

2232

uses 2.7 ohms of resistance with a 4700 uF capacitor)

tom jobe....

The charging current is roughly 1/2 x 19 = 9.5 V peak. The resistor is 3

ohms,

so the current is limited to about 3 amps.

I = C dV/dT, soI/C = dV/dTso3 / 0.0047 = 14,100V/sec = 0.014

V/uS

so in 25 uS (1/2 period)it rises 25 x 0.014 = 0.35 V

that's about 1V in 3 cycles or 10 V in 30 cycles, which seems reasonable.

Best,

-John

PS I'm not trying to design anything here, I'm just trying to learn the

basics with small projects like this, and I thank all of you for helping

me.

----- Original Message -----

From: "J Forster" <jfor@...>

To: <TekScopes2@...>

Sent: Saturday, February 02, 2008 7:18 PM

Subject: Re: [TekScopes2] Optional fan for the 2235 or USM-488?

[snip]In one of my books it talked about how you need about .5

volts

of ripple in the first stage of a power supply to make sure both diodes

are

actually working, and I only had 60 mV.

It is good fun learning about electronics with this fine Tektronix

gear!

tom jobe...

No. If you look at the waveform to a capacitor input filter on a 60 Hz

line, you should see an increased voltage bump every 1/120th of a

second. If

you see one every 1/60th of a second, it's a half wave rectification.

The amplitude depends on the C size and load current:

I = CdV/dTso dV = I / C dT

dV is roughly the P-P voltage of the ripple

dT is the interval between charging pulses(1/60 or 1/120 second)

SO:

dVincreases w/ current drawand decreases w/ capacitor size.

Best,

-John

Hello Arthur,

Yes, the diodes I used came from a junk 2235 mainboard, and they were

rectifier diodes from one of the secondary supplies. I checked them on the

curve tracer and they seemed to be identical. You are right about getting

almost peak voltage across the large 1000 uF cap.

Thanks for your help!

tom jobe...






On 8/4/2018 7:47 PM, Roy Morgan wrote:

Tom,

I would want to do this fan modification to my 2215. A modest amount of cooling can help electronics life span a lot.

So I’d be most grateful for the pictures you have.

Thanks in advance.


Roy Morgan
k1lky68@... <mailto:k1lky68@...>

On Aug 4, 2018, at 8:08 PM, tom jobe <tomjobe@... <mailto:tomjobe@...>> wrote:

Hello,
Do not forget that scopes such as the 2215A and the 2235 have the fan drive circuitry you need already printed on the mainboard,...
I have some photos of the above modifications that I can send to anyone directly on request.

BTW...have a read on this snubbing guide, especially concerning contact bounce which exists no matter how 'crisp' the switch action is.

On Sun, Aug 5, 2018 at 11:50 AM, Vincent Trouilliez wrote:

On Sun, Aug 5, 2018 at 12:38 AM, satbeginner wrote:

I had a quick look on the 2232 schematics
The 2232 has a full wave rectifier (CR965 & CR967 coming from pin 20 and 22

of the transformer)

That's true for the early units, which drive the fan at 8.6V via the a middle
tap and the two +/- 8.6 rails.
But I see there was a different wiring for later 2232's : this time it's
running at a slightly higher voltage : 10V. It's across the +/- 5V rails and
doesn't use the middle tap/ground, hence a 4 diode bridge.

I don't know why they changed the wiring. Either they felt they needed a bit
more voltage and/or they decided that running a noise factory such as a fan,
from the 8.6V rails that power the sensitive analog stuff, was far from ideal,
even though the fan is wired upstream of the LC filtering networks.

and a capacitor C965 and a current limiting resistor R965 installed to power

the additional fan.

At least this bit is common to both designs...

Based on the empty room on my 2215A PCB it looks like they only planned a

half

wave rectifier?
There is room for 1 diode coming from pin 20 of the transformer, 1 capacitor
and 1 resistor and a 2-pin connector for the fan.

OK ! So yet another wiring arrangement ! LOL

So in short... in all cases they put that 2.7ohm resistor in series, and that
big 4700uF cap across the fan... but as far as voltage source goes, you pretty
much tap the secondaries in any way that you see fit, so long as you tap
straight from the transformer pins so that the crap/electrical noise from the
fan can still be filtered downstream by the LC PI networks, and keep the
delicate analog stuff of the scope happy.

Well that's how I understand it so far...

Will try this,

Yes, keep us posted, and take some waveforms and measurements to see how this
crude single wave rectification performs... and of course measure ripple
before/after the mod, to make sure it's still within specs ^^


Regards,

Vincent Trouilliez

Un saludo,

Leo

Hi Vincent,

I took some pictures of the empty spaces on my PCB and how I filled them with the three components and the connector.
The pictures and the additional schematics are added to my album /g/TekScopes/album?id=64919

For now I cannot measure any (extra) ripple because I do not have the fan yet.. :-)

To be continued,

Leo

On Sun, Aug 5, 2018 at 12:38 AM, satbeginner wrote:

I had a quick look on the 2232 schematics
The 2232 has a full wave rectifier (CR965 & CR967 coming from pin 20 and 22 of the transformer)

That's true for the early units, which drive the fan at 8.6V via the a middle tap and the two +/- 8.6 rails.
But I see there was a different wiring for later 2232's : this time it's running at a slightly higher voltage : 10V. It's across the +/- 5V rails and doesn't use the middle tap/ground, hence a 4 diode bridge.

I don't know why they changed the wiring. Either they felt they needed a bit more voltage and/or they decided that running a noise factory such as a fan, from the 8.6V rails that power the sensitive analog stuff, was far from ideal, even though the fan is wired upstream of the LC filtering networks.

and a capacitor C965 and a current limiting resistor R965 installed to power the additional fan.

At least this bit is common to both designs...

Based on the empty room on my 2215A PCB it looks like they only planned a half
wave rectifier?
There is room for 1 diode coming from pin 20 of the transformer, 1 capacitor
and 1 resistor and a 2-pin connector for the fan.

OK ! So yet another wiring arrangement ! LOL

So in short... in all cases they put that 2.7ohm resistor in series, and that big 4700uF cap across the fan... but as far as voltage source goes, you pretty much tap the secondaries in any way that you see fit, so long as you tap straight from the transformer pins so that the crap/electrical noise from the fan can still be filtered downstream by the LC PI networks, and keep the delicate analog stuff of the scope happy.

Well that's how I understand it so far...

Will try this,

Yes, keep us posted, and take some waveforms and measurements to see how this crude single wave rectification performs... and of course measure ripple before/after the mod, to make sure it's still within specs ^^


Regards,

Vincent Trouilliez

Un saludo,

Leo

On Sat, Aug 4, 2018 at 07:56 PM, Roy Morgan wrote:

Vincent,

This is certainly a mod I want to do to my 2215. A modest amount of air improves the life of electronics a lot.
I’d be grateful to get the pictures you have, with my thanks.

Roy Morgan

Hi Roy,

It was Tom Jobe (which you quoted) who offered this info, not me ;-P

I am interested in it too ! :-)

As for posting Photos, I think it's much better to put them in a clearly/explicitly named, dedicated album, because of it is of general interest for sure.


Vincent Trouilliez

Hello,

I had a quick look on the 2232 schematics and my 2215A board layout:
The 2232 has a full wave rectifier (CR965 & CR967 coming from pin 20 and 22 of the transformer) and a capacitor C965 and a current limiting resistor R965 installed to power the additional fan.

Based on the empty room on my 2215A PCB it looks like they only planned a half wave rectifier?
There is room for 1 diode coming from pin 20 of the transformer, 1 capacitor and 1 resistor and a 2-pin connector for the fan.

Will try this,

Un saludo,

Leo

Hi,

I for sure want to add a fan too!

The album I made is open for the group, so you can simply add the pictures here: /g/TekScopes/album?id=64919

Un saludo,

Leo

Tom,

I recommend posting on groups.io

George

BTW, you owe me a lunch time visit...

Vincent,

This is certainly a mod I want to do to my 2215. A modest amount of air improves the life of electronics a lot.

I’d be grateful to get the pictures you have, with my thanks.


Roy Morgan
k1lky68@... <mailto:k1lky68@...>