The Museum of HP Calculators

HP Forum Archive 15

[ Return to Index | Top of Index ]

Successful HP-11C speedup
Message #1 Posted by Karl Schneider on 18 Sept 2005, 10:25 p.m.

Fans of the HP-1*C "Voyager" models --

I followed the instructions posted in the article "HP Voyager Calculator Speedup" by Ken Sumrall and (later) Eric Smith to accelerate an 11C that I recently bought. The article is available at

http://www.brouhaha.com/~eric/hpcalc/voyager_speedup/

The calculator is a 1985 model (S/N 2508Annnnn) with the older 1LF5 CPU. I soldered a small 33 uH (microhenry) inductor to the same two terminals on the circuit board just below the battery compartment, where the calculator's 180 uF inductor underneath the board is soldered. The additional inductor fits between the case and the circuit board, and is electrically paralleled with the 180 uH inductor.

The modification reduces the inductance of the parallel LC circuit, increasing its resonant frequency and the effective clock speed, as described in the article.

Acceleration of processing is as described in the article -- about 2.5 times as fast as an unmodified 1986 11C (S/N 2624Annnnn), corresponding to the higher clock frequency. I have yet to see any malfunction or incorrect calculation result.

I bought the small 33 uH and 22 uH inductors at Norvac Electronics in Beaverton, Oregon for $0.60 each. (The 47 uH and 28 uH inductors recommended by Ken were not available.) The one Radio Shack store I asked carried only a rather large 100 uH inductor, but not these compact ones of lower inductance.

Speedup would be particularly useful on the 15C (for Solve and Integrate), but is a welcome enhancement for any Voyager. I just hope that there aren't any unwelcome side effects, besides the tolerable ones listed in the article.

The 11C I modified has a piece of the metal backing plate and one of the chassis-grounding springs missing, but the calc exhibits no apparent malfunction. Just curious: Is the second spring needed to connect one part of the board to another, or are two springs used only for redundancy?

-- KS

Edited: 23 Sept 2005, 12:28 a.m. after one or more responses were posted

      
Re: Successful HP-11C speedup
Message #2 Posted by Matthias Wehrli on 19 Sept 2005, 1:00 a.m.,
in response to message #1 by Karl Schneider

Hi Karl Did you take some pictures of your modification? Can you let us see them?

Matthias

            
Re: Successful HP-11C speedup
Message #3 Posted by marais on 19 Sept 2005, 2:43 a.m.,
in response to message #2 by Matthias Wehrli

Well, one unwelcome side effect will be an increased power consumption, approx. by a factor of 2.5

Andreas

                  
Re: Successful HP-11C speedup
Message #4 Posted by Eric Smith on 20 Sept 2005, 7:27 p.m.,
in response to message #3 by marais

Only while it's actually computing. When it's not running or processing a keystroke, the CPU goes into a very low power sleep mode, and the oscillator doesn't run.

So with the mod, you get 2.5x the performance for 2.5x the power, thus using the same amount of power for your calculations.

      
Re: Successful HP-11C speedup
Message #5 Posted by Gerson W. Barbosa on 19 Sept 2005, 9:31 a.m.,
in response to message #1 by Karl Schneider

Hi Karl,

Congratulations for your success! My newer (or should I say less old?) 15C has the 1LM2 CPU. In theory, it can be speeded up up to 5.3 times. I'll try a 2 or 2.5 speedup. The drawback is that batteries may last only 3 years rather than 6 or 8, under the way I use it. The article says something about a narrower operating temperature range. No problem! According to the manual, the operating range is 0C to 55. Here I estimate we have indoor temperatures ranging from 10 to 30 all the year round. There is only one visible capacitor next to the battery compartment. Is this as simple as soldering the inductor in parallel with it?

Thanks,

Gerson.

            
Re: Successful HP-11C speedup
Message #6 Posted by don wallace on 19 Sept 2005, 9:43 a.m.,
in response to message #5 by Gerson W. Barbosa

Congrats Karl on your successful mod.

I imagine you will not have any problems at all. At extremes of temperature or at low battery volatge you may possibly have problems with the faster clocking, but these will, if they occur, result in either the calc not turning on, not turning off, or very strange and dramatic malfunctions as the memory or display does not respond to the faster logic operations on the buss. I beleive if this were to ever occur the worst that might happen is that the machine memory will get scrambled and the processor will do the 11C equivalent of a "memory lost" (I only have a 41c and cv). In that case, just increasing the inductor value a bit will slow it down a bit and get you reliable operation again. As you probably know, The 11c uses the same processor as the 41c. If you want any advice on how the processor works, feel free to ask me. I have been doing a bit of tinkering with the coconut... (and a lot of reading... ;-)

Don W

      
Answers: Successful HP-11C speedup
Message #7 Posted by Karl Schneider on 19 Sept 2005, 11:00 p.m.,
in response to message #1 by Karl Schneider

Quote:
Did you take some pictures of your modification? Can you let us see them?

There is only one visible capacitor next to the battery compartment. Is this as simple as soldering the inductor in parallel with it?


I decided against taking pictures because my "simple" digital camera is a hassle and it discharges its batteries. I wanted to make sure the calc worked properly before inconveniencing myself by taking photos.

Of course, I can always open it up again and take them.

I was hoping that the Voyager internal photos at this website would show clearly where the modification was made. Unfortunately, the photo of the 15C (http://www.hpmuseum.org/10cint.jpg) looks different from the 11C I modified. The prominent tubular object with colored bands looks like the 180 uH inductor that is under the circuit board in my 11C.

The article says that the stock 180 pF capacitor and 180 uH inductor form a parallel LC circuit. So, the additional inductor could be soldered in parallel with either one, presumably. I soldered the new inductor in parallel with the existing inductor, because that was straightforward.

I also have an HP-12C made in 1990 that I can try to "enhance".

-- KS

Edited: 23 Sept 2005, 12:28 a.m. after one or more responses were posted

            
Re: Answers: Successful HP-11C speedup
Message #8 Posted by don wallace on 19 Sept 2005, 11:25 p.m.,
in response to message #7 by Karl Schneider

Hi Karl.

I think you mean 180 pF capacitor.

For general info for thread readers, inductors in parallel drop the value (desired goal) and for capacitors it's in SERIES to drop the value.

In the 41C series the parts are: 82 uH and 150 pF. The resonant frequency of that is around 1440 kHz. Logic in the processor does a divide by four to get the two system clocks which run at about 360kHz. It interested me to read that the 11 and 12 use the same processor as the 41. But the 41 has a voltage regulator and a 6 volt buss so they (can reliably) clock it faster.

(pF = picofarad, one millionth of a uF, stating the obvious for the casual reader)

Pardon me for sticking my two cents worth in...

DW

Edited: 20 Sept 2005, 1:30 a.m.

                  
Thank you, Don...
Message #9 Posted by Karl Schneider on 21 Sept 2005, 12:52 a.m.,
in response to message #8 by don wallace

Don Wallace stated,

Quote:
Hi Karl.

I think you mean 180 pF capacitor.

(pF = picofarad, one millionth of a uF.)

Pardon me for sticking my two cents worth in...


Mea culpa! You are right -- no apologies necessary. I failed to notice in Ken's article what was actually there. With 180 uF, the resonant frequency on the 11C is 884.19 Hz, not 884.19 kHz.

Also thanks for posting the helpful precautions. Next time I'll follow them...

-- KS

Edited: 21 Sept 2005, 1:54 a.m.

            
Re: Answers: Successful HP-11C speedup
Message #10 Posted by Matthias Wehrli on 20 Sept 2005, 10:29 a.m.,
in response to message #7 by Karl Schneider

Hi Karl and others

I woul dbe glad to get a picture of the HP-11C speedup anyway. Not to verify your work, but to see, how to do. Im not exactly shure, if I understood the instructions correctly and I do not want to kill a HP-11C... If yo do not want to send the picture to the public, you can also send it to me personaly...

Matthias

                  
Re: Answers: Successful HP-11C speedup
Message #11 Posted by don wallace on 20 Sept 2005, 5:30 p.m.,
in response to message #10 by Matthias Wehrli

Hi Mathias.

The main thing to be wary of is causing damage due to static electricity (electrostatic damage or esd in trade terms).

The simplest way to avoid this is to touch the soldering iron (plugged into a grounded three pin wall outlet) briefly on an electrical ground point in the calculator electrical system. The copper of a piezoelectric buzzer is a good ground point. Do not apply heat, just make contact.

This brings the calculator electrically to ground in a safe way.

Having done that the next thing is to be sure of the timing capacitor and timing inductor identity and to be careful in not damaging the pcb.

I find it is better to put anther capacitor in series with the timing cap. as you can just cut one wire on the cap itself and put the extra capacitor where the cut was made. I choose capacitors as they are less bulky than inductors.

If you want to choose the inductor, it looks like a fairly large bluish power resistor with colour code bands. And using a multimeter it looks like a short circuit reading very low resistance of say one ohm.

There is a photo somwhere on my site (on the 41).

DW

                        
Re: Answers: Successful HP-11C speedup
Message #12 Posted by Gerson W. Barbosa on 21 Sept 2005, 3:28 p.m.,
in response to message #11 by don wallace

What if just replacing the capacitor instead of adding another in series? That might be a better option for earlier boards. On newer boards, it may be easier just to solder another inductor in parallel with the existing one. In this case, the terminals should not be overheated as the existing weld may melt under the other side of the board. I have to open up the calculator to see if there's room for another inductor. In the table below (DECIMAL POINT IS COMMA) are listed both options (I am not sure if all capacitors are easily available). Notice that the mutual inductance between inductors have not been taken into account in the calculations. The table quite matches one of Ken's while the other is slightly different, which makes me wonder he may have soldered the new inductor below the existing one in the older board, thus minimizing the influence of the mutual inductance, and side by side on the newer board. Just a guess though. Please correct me if I am wrong.

Gerson.

-----------------------------------------------------------------

   

L (uH) C (pF) f (kHz)* speedup factor

original 180 180 884,19 1,00 LC circuit

180 180 1250,44 1,41 100 180 1479,54 1,67 82 180 1580,49 1,79 inductor 47 180 1943,17 2,20 in 33 180 2246,37 2,54 parallel 22 180 2679,24 3,03 15 180 3188,01 3,61 12 180 3536,78 4,00 10 180 3854,11 4,36 8,2 180 4235,95 4,79

180 100 1186,27 1,34 180 82 1310,02 1,48 capacitor 180 47 1730,35 1,96 replacement 180 33 2065,03 2,34 180 22 2529,14 2,86 180 15 3062,94 3,46 180 12 3424,47 3,87 180 10 3751,32 4,24

* not actual clock frequencies

Edited: 21 Sept 2005, 3:35 p.m.

                              
Re: Answers: Successful HP-11C speedup
Message #13 Posted by don wallace on 21 Sept 2005, 5:16 p.m.,
in response to message #12 by Gerson W. Barbosa

Hi Gerson.

Yes, just replacing the capacitor is fine.

Mutual inductance effect is not a significant factor. The coupling factor would be very low... But at least you are thinking about things, which is good.

There are other factors which limit the top speed of the clock. The active components in the chip itself start to limit things as you go up in speed. So top speed has to be found from experiment. And this figure will vary a bit from calc to calc because they are all individuals... ;-)

Process variations at the factory mean each chip will have slight differences in it's absolute maximum speed ratings.

DW

                                    
Successful HP-15C speedup!
Message #14 Posted by Gerson W. Barbosa on 21 Sept 2005, 7:31 p.m.,
in response to message #13 by don wallace

Hi Don,

I've just done a speedup in one of my 15C's. In this case, it was easier to add an inductor. Using a 33uH inductor I obtained a 2.49x speedup. It ran a benchmark routine in 9.75 s (mean of three measurements). Previously it had taken 24.3 s to run the same test. I will provide 'before' and 'after' pictures somewhere else in the thread later, in case other people are interested.

Cheers,

Gerson.

                                          
Re: Successful HP-15C speedup!
Message #15 Posted by don wallace on 21 Sept 2005, 7:47 p.m.,
in response to message #14 by Gerson W. Barbosa

Cool!

DW

      
Successful HP-15C speedup! (Pictures)
Message #16 Posted by Gerson W. Barbosa on 21 Sept 2005, 9:06 p.m.,
in response to message #1 by Karl Schneider

I have just sped up my 2547B2728 15C (2.49x, using a 33uH inductor - see my reply to Don for other options). Thanks Karl, for having guts to try, thanks Eric for keeping Ken's article and my curiosity (you may remember I noticed Tony's 2-speed 15C among his numerous calculators of his collection).

When I left work at 05:00 PM, instead of going home as usual, I decided to go for the inductors. Unfortunately the ones I could find here were just too large to fit inside the calculator. So, I removed its outer blue cap, as you can see in the picture below. When I opened up the calculator after taking the batteries out I looked for two low-impedance points close to the battery compartment. I measured 6.4 ohm between the first two points I tried, those circled in red. I had to presume it was the LC circuit, because I could not see underneath the board. As the soldering iron I have at home is not grounded, I removed it from the outlet when soldering as a precaution. The green wire in the second picture was soldered to one end of the inductor because it was too short.

Before one decides to try, I would suggest a careful checking in the board as printed circuit board layouts may vary a little. It is advisable to follow all safety recommendations listed throughout this thread. The risk of permanently damaging the calculator while doing this modification should not be overlooked.

Regards,

Gerson.

http://fotos.aol.com.br/adAlbumFotoThumbList.aspx?Album=27187

For access to the pictures enter 1234 in Senha (Password) 
Then click on Enviar (Send)
Click on Salvar (Save) above pictures to save actual size photos

Edited: 30 Sept 2005, 9:29 p.m.


[ Return to Index | Top of Index ]

Go back to the main exhibit hall