My Surprisingly Fast 41CX

[akpoff]

In another thread I mentioned this 41CX shows signs of prior surgery, and now I know more.

This 41CX is fast, not 41CL fast, but probably LC-circuit-fiddling faster.

I initially noticed pressing keys would often show the key name briefly before executing the function, where my other 41CX would not. Yesterday I happened to XEQ BEEP and got the chipmunk version, which led me to comparing longer-running functions like FACT 40 against the other 41CX. It's noticeably quicker, nothing like a DM42, but quicker compared to stock.

The only thing left to do, then, was to figure out which approach they took. In message 3 of this thread Theo Kas mentions upgrading his calculator back in the day by replacing the 140pF capacitor with a 22pF cap. He also notes common values range from 22pF to 33pF.

A closer look at my circuit board shows a CM4ED 0220J0 capacitor where the 140pF capacitor normally sits, hidden below the piezo-electric buzzer.

   

Unlike some of the other options, mine isn't switchable. It's turbo all the time. I tested a card reader and wand with it. So far no ill effects. In fact, other than flashes of key names when I press keys too slowly, and chipmunk sound effects it works just fine.

Mystery solved. My little 41CX is full of surprises.

Edit: Correct Clonix with 41CL. See below.

Cheers,

--Aaron

[Sylvain Cote]

(01-25-2023 07:03 PM)akpoff Wrote:  This 41CX is fast, not Clonix fast, but probably LC-circuit-fiddling faster.

I am happy that your HP-41 is fast, but saying that it is not Clonix fast make no sense.
For sure, Diego's modules add a lot of functionalities, but changing the HP-41 speed, they do not!

[akpoff]

(01-25-2023 07:42 PM)Sylvain Cote Wrote:  I am happy that your HP-41 is fast, but saying that it is not Clonix fast make no sense.
For sure, Diego's modules add a lot of functionalities, but changing the HP-41 speed, they do not!

That was an autopilot typo. I had in mind the 41CL circuit board upgrade but typed out Clonix. Thanks for catching that.

[Garth Wilson]

I'm glad it seems to work fine.  In the mid- or late-1980's, EduCalc was selling a speed-up kit which was supposed to make the 41 70% faster IIRC.  I bought one, but I read all the stuff before beginning the conversion, and it said you shouldn't do it if you use the HP-IL or combined modules.  I had both, and used them constantly, so I sent it back.

[akpoff]

(01-25-2023 07:57 PM)Garth Wilson Wrote:  I'm glad it seems to work fine.  In the mid- or late-1980's, EduCalc was selling a speed-up kit which was supposed to make the 41 70% faster IIRC.  I bought one, but I read all the stuff before beginning the conversion, and it said you shouldn't do it if you use the HP-IL or combined modules.  I had both, and used them constantly, so I sent it back.

Hi Garth,
I saw your comment about that in another thread. I have the HP-IL module and a few HP-IL devices. The quickest to test will be the 82162A Thermal Printer.

Do you recall the specific concerns with HP-IL? Timing would be an obvious possibility, though the real-time clock seems to be unaffected.

As I understand the 41CL documentation, Monte's approach is way more refined in that it doesn't blindly speed-up everything, but only where it make sense. Of course he has the advantage of having designed his solution rather than having to work within the constraints of an existing circuit design. (Well, the 41CL still has some constraints to work around. )

Thanks for the note!

--Aaron

[Garth Wilson]

akpoff, what the papers that came with the speed-up kit seemed to be saying was that signals at the ports would not give adequate setup and hold times for reliable operation with the IL and combined modules.  I had already bought the HP-IL and Extended-I/O combined module so the two would only take a single port, and I was using this constantly (and not just for the ThinkJet printer but also going through the HP82169A HPIL-to-IEEE488 adapter to several pieces of instrumentation on the workbench).  (Later I got a double XMem combined module also.)  I didn't really want to try it and find out, possibly later, that it would fail under certain conditions.  There was already the risk of damaging something like the screw posts merely from opening it, and then also of not being able to return the kit after installing it and finding that I couldn't use it.

When Monte originally announced the 41CL, I hadn't understood what you were saying about still keeping the slower timings for IL and other modules.  What he did was outstanding though.